How Shall We Measure Progress Over the Past Three Years?

For a dozen years, my career has been centered on Internet search, dynamic content and the deep Web. For the past few years, I have been somewhat obsessed by two topics.

The first topic, a conviction really, is that implicit structure needs to be extracted from Web content to enable it to be disambiguated, organized, shared and re-purposed. The second topic, more an open question as a former academic married to a professor, is what might replace editorial selections and peer review to establish the authoritativeness of content. These topics naturally steer one to the semantic Web.

A Millennial Perspective

The semantic Web, by whatever name it comes to be called, is an inevitability. History tells us that as information content grows, so do the mechanisms for organizing and managing it. Over human history, innovations such as writing systems, alphabetization, pagination, tables of contents, indexes, concordances, reference look-ups, classification systems, tables, figures, and statistics have emerged in parallel with content growth [19].

When the Lycos search engine, one of the first profitable Internet ventures, was publicly released in 1994, it indexed a mere 54,000 pages [1]. When Google wowed us with its page-ranking algorithm in 1998, it soon replaced my then favorite search engine, AltaVista. Now, tens of billions of indexed documents later, I often find Google’s results to be overwhelming dross — unfortunately true again for all of the major search engines. Faceted browsing, vertical search, and Web 2.0’s tagging and folksonomies demonstrate humanity’s natural penchant to fight this entropy, efforts that will next continue with the semantic Web and then mechanisms unforeseen to manage the chaos of burgeoning content.

An awful lot of hot air has been expelled over the false dichotomy of whether the semantic Web will fail or is on the verge of nirvana. Arguments extend from the epistemological versus ontological (classically defined) to Web 3.0 versus SemWeb or Web services (WS*) versus REST (Representational State Transfer). My RSS feed reader points to at least one such dust up every week.

Some set the difficulties of resolving semantic heterogeneities as absolutes, leading to an illogical and false rejection of semantic Web objectives. In contrast, some advocates set equally divisive arguments for semantic Web purity by insisting on formal ontologies and descriptive logics. Meanwhile, studied leaks about “stealth” semantic Web ventures mean you should grab your wallet while simultaneously shaking your head.

A Decades-Long Perspective

My mental image of the semantic Web is a road from here to some achievable destination — say, Detroit. Parts of the road are well paved; indeed, portions are already superhighways with controlled on-ramps and off-ramps. Other portions are two lanes, some with way too many traffic lights and some with dangerous intersections. A few small portions remain unpaved gravel and rough going.

Wreck in Nebraska during the 1919 Transcontinental Motor Convoy

A lack of perspective makes things appear either too close or too far away. The automobile isn’t yet a century old as a mass-produced item. It wasn’t until 1919 that the US Army Transcontinental Motor Convoy made the first automobile trip across the United States.

The 3,200 mile route roughly followed today’s Lincoln Highway, US 30, from Washington, D.C. to San Francisco. The convoy took 62 days and 250 recorded accidents to complete the trip (see figure), half on dirt roads at an average speed of 6 miles per hour. A tank officer on that trip later observed Germany’s autobahns during World War II. When he subsequently became President Dwight D. Eisenhower, he proposed and then signed the Interstate Highway Act.

That was 50 years ago. Today, the US is crisscrossed with 50,000 miles of interstates, which have completely remade the nation’s economy and culture [2].

Today’s Perspective

Like the interstate system in its early years, today’s semantic Web lets you link together a complete trip, but the going isn’t as smooth or as fast as it could be. Nevertheless, making the trip is doable and keeps improving day by day, month by month.

My view of what’s required to smooth the road begins with extracting structure and meaningful information according to understandable schema from mostly uncharacterized content. Then we store the now-structured content as RDF triples that can be further managed and manipulated at scale. By necessity, the journey embraces tools and requirements that, individually, might not constitute semantic Web technology as some strictly define it. These tools and requirements are nonetheless integral to reaching the destination. We are well into that journey’s first leg, what I and others are calling the structured Web.

For the past six months or so I have been researching and assembling as many semantic Web and related tools as I can find [3]. That Sweet Tools listing now exceeds 500 tools [4] (with its presentation using the nifty lightweight Exhibit publication system from MIT’s Simile program [5]). I’ve come to understand the importance of many ancillary tool sets to the entire semantic Web highway, such as natural language processing and information extraction. I’ve also found new categories of pragmatic tools that embody semantic Web and data mediation processes but don’t label themselves as such.

In its entirety, the Sweet Tools listing provides a pretty good picture of the semantic Web’s state. It’s a surprisingly robust picture — though with some notable potholes — and includes impressive open source options in all categories. Content publishing, indexing, and retrieval at massive scales are largely solved problems. We also have the infrastructure, languages, and (yes!) standards for tying this content together meaningfully at the data and object levels.

I also think a degree of consensus has emerged on RDF as the canonical data model for semantic information. RDF triple stores are rapidly improving toward industrial strength, and RESTful designs enable massive scalability, as terabyte- and petabyte-scale full-text indexes prove.

Powerful and flexible middleware options, such as those from OpenLink [6], can transform and integrate diverse file formats with a variety of back ends. The World Wide Web Consortium’s GRDDL standard [7] and related tools, plus various “RDF-izers” from Massachusetts Institute of Technology and elsewhere [8], largely provide the conversion infrastructure for getting Web data into that canonical RDF form. Sure, some of these converters are still research-grade, but getting them to operational capabilities at scale now appears trivial.

Things start getting shakier when trying to structure information into a semantic formalism. Controlled vocabularies and ontologies range broadly and remain a contentious area. Publishers and authors perhaps have too many choices: from straight Atom or RSS feeds and feeds with tags to informal folksonomies and then Outline Processor Markup Language [9] or microformats [10]. From there, the formalism increases further to include the standard RDF ontologies such as SIOC (Semantically-Interlinked Online Communities), SKOS (Simple Knowledge Organizing System), DOAP (Description of a Project), and FOAF (Friend of a Friend) [11] and the still greater formalism of OWL’s various dialects [12].

Arguing which of these is the theoretical best method is doomed to failure, except possibly in a bounded enterprise environment. We live in the real world, where multiple options will always have their advocates and their applications.

All of us should welcome whatever structure we can add to our information base, no matter where it comes from or how it’s done. The sooner we can embrace content in any of these formats and convert it into canonical RDF form, we can then move on to needed developments in semantic mediation, some of the roughest road on the journey.

Potholes on the Semantic Highway

Semantic mediation requires appropriate structured content. Many potholes on the road to the semantic Web exist because the content lacks structured markup; others arise because existing structure requires transformation. We need improved ways to address both problems. We also need more intuitive means for applying schema to structure. Some have referred to these issues as “who pays the tax.”

Recent experience with social software and collaboration proves that a portion of the Internet user community is willing to tag and characterize content. Furthermore, we can readily leverage that resulting structure, and free riders are welcomed. The real pothole is the lack of easy — even fun — data extractors and “structurizers.” But we’re tantalizingly close.

Tools such as Solvent and Sifter from MIT’s Simile program [13] and Marmite from Carnegie Mellon University [14] are showing the way to match DOM (document object model) inspectors with automated structure extractors. DBpedia, the alpha version of Freebase, and System One now provide large-scale, open Web data sets in RDF [15], including all of Wikipedia. Browser extensions such as Zotero [16] are showing how to integrate structure management into acceptable user interfaces, as are services such as Zoominfo [17]. Yet we still lack easy means to design the differing structures suitable for a plenitude of destinations.

Amazingly, a compelling road map for how all these pieces could truly fit together is also incomplete. How do we actually get from here to Detroit? Within specific components, architectural understandings are sometimes OK (although documentation is usually awful for open source projects, as most of the current tools are). Until our community better documents that vision, attracting new contributors will be needlessly slower, thus delaying the benefits of network effects.

So, let’s create a road map and get on with paving the gaps and filling the potholes. It’s not a matter of standards or technology — we have those in abundance. Let’s stop the silly squabbles and commit to the journey in earnest. The structured Web’s ability to reach Hyperland [18], Douglas Adam’s prescient 1990 forecast of the semantic Web, now looks to be no further away than Detroit.

This Friday brown bag leftover was first placed into the AI3 refrigerator about three years ago on May 3, 2007.  The piece was my answer to a request by Jim Hendler to pen some thoughts on the semantic Web, based on I believe what he thought might be a pragmatic perspective combining Internet business with Web science. The formal piece appeared as a guest editorial in the May/June 2007 issue of IEEE Intelligent Systems. What appears above is unaltered from my original posting (aside from some minor formatting clean-up and — sorry to say — some of the projects are now defunct).

A Reprise AI3 Post from Four Years Ago

In 2002 Joel Mokyr, an economic historian from Northwestern University, wrote a book that should be read by anyone interested in knowledge and its role in economic growth. The Gifts of Athena : Historical Origins of the Knowledge Economy is a sweeping and comprehensive account of the period from 1760 (in what Mokyr calls the “Industrial Enlightenment”) through the Industrial Revolution beginning roughly in 1820 and then continuing through the end of the 19th century.

The book (and related expansions by Mokyr available as separate PDFs on the Internet) should be considered as the definitive reference on this topic to date. The book contains 40 pages of references to all of the leading papers and writers on diverse technologies from mining to manufacturing to health and the household. The scope of subject coverage, granted mostly focused on western Europe and America, is truly impressive.

Mokyr deals with ‘useful knowledge,’ as he acknowledges Simon Kuznets‘ phrase. Mokyr argues that the growth of recent centuries was driven by the accumulation of knowledge and the declining costs of access to it. Mokyr helps to break past logjams that have attempted to link single factors such as the growth in science or the growth in certain technologies (such as the steam engine or electricity) as the key drivers of the massive increases in economic growth that coincided with the era now known as the Industrial Revolution.

Mokyr cracks some of these prior impasses by picking up on ideas first articulated through Michael Polanyi’s “tacit knowing” (among other recent philosophers interested in the nature and definition of knowledge). Mokyr’s own schema posits propositional knowledge, which he defines as the science, beliefs or the epistemic base of knowledge, which he labels omega (Ω), in combination with prescriptive knowledge, which are the techniques (”recipes”), and which he also labels lambda (λ). Mokyr notes that an addition to omega (Ω) is a discovery, an addition to lambda (λ) is an invention.

One of Mokyr’s key points is that both knowledge types reinforce one another and, of course, the Industrial Revolution was a period of unprecedented growth in such knowledge. Another key point, easily overlooked when “discoveries” are seemingly more noteworthy, is that techniques and practical applications of knowledge can provide a multiplier effect and are equivalently important. For example, in addition to his main case studies of the factory, health and the household, he says:

The inventions of writing, paper, and printing not only greatly reduced access costs but also materially
affected human cognition, including the way people thought about their environment.

Mokyr also correctly notes how the accumulation of knowledge in science and the epistemic base promotes productivity and more still-more efficient discovery mechanisms:

The range of experimentation possibilities that needs to be searched over is far larger if the searcher knows nothing about the natural principles at work. To paraphrase Pasteur’s famous aphorism once more, fortune may sometimes favor unprepared minds, but only for a short while. It is in this respect that the width of the epistemic base makes the big difference.

In my own opinion, I think Mokyr starts to get closer to the mark when he discusses knowledge “storage”, access costs and multiplier effects from basic knowledge-based technologies or techniques. Like some other recent writers, he also tries to find analogies with evolutionary biology. For example:

Much like DNA, useful knowledge does not exist by itself; it has to be “carried” by people or in storage
devices. Unlike DNA, however, carriers can acquire and shed knowledge so that the selection process is quite different. This difference raises the question of how it is transmitted over time, and whether it can actually shrink as well as expand.

One of the real advantages of this book is to move forward a re-think of the “great man” or “great event” approach to history. There are indeed complicated forces at work. I think Mokyr summarizes well this transition when he states:

A century ago, historians of technology felt that individual inventors were the main actors that brought about
the Industrial Revolution. Such heroic interpretations were discarded in favor of views that emphasized deeper economic and social factors such as institutions, incentives, demand, and factor prices. It seems, however, that the crucial elements were neither brilliant individuals nor the impersonal forces governing the masses, but a small group of at most a few thousand peopled who formed a creative community based on the exchange of knowledge. Engineers, mechanics, chemists, physicians, and natural philosophers formed circles in which access to knowledge was the primary objective. Paired with the appreciation that such knowledge could be the base of ever-expanding prosperity, these elite networks were indispensible, even if individual members were not. Theories that link education and human capital of technological progress need to stress the importance of these small creative communities jointly with wider phenomena such as literacy rates and universal schooling.

There is so much to like and to be impressed with this book and even later Mokyr writings. My two criticisms are that, first, I found the pseudo-science of his knowledge labels confusing (I kept having to mentally translate the omega symbol) and I disliked the naming distinctions between propositional and prescriptive, even though I think the concepts are spot on.

My second criticism, a more major one, is that Mokyr notes, but does not adequately pursue, “In the decades after 1815, a veritable explosion of technical literature took place. Comprehensive technical compendia appeared in every industrial field.” Statements such as these, and there are many in the book, hint at perhaps some fundamental drivers.

Mokyr has provided the raw grist for answering his starting question of why such massive economic growth occurred in conjunction with the era of the Industrial Revolution. He has made many insights and posited new factors to explain this salutary discontinuity from all prior human history. But, in this reviewer’s opinion, he still leaves the why tantalizingly close but still unanswered. The fixity of information and growing storehouses because of declining production and access costs remain too poorly explored.

This Friday brown bag leftover was first placed into the AI3 refrigerator about four years ago on July 6, 2006. It was part of a series of book reviews I was doing at that time getting at the importance of bulk paper production as a key enabler of economic growth. No changes have been made to the original posting.

In earlier posts, I described the significant progress in climbing the data federation pyramid, today’s evolution in emphasis to the semantic Web, and the 40 or so sources of semantic heterogeneity. We now transition to an overview of how one goes about providing these semantics and resolving these heterogeneities.

Why the Need for Tools and Automation?

In an excellent recent overview of semantic Web progress, Paul Warren points out:[1]

Although knowledge workers no doubt believe in the value of annotating their documents, the pressure to create metadata isn’t present. In fact, the pressure of time will work in a counter direction. Annotation’s benefits accrue to other workers; the knowledge creator only benefits if a community of knowledge workers abides by the same rules. . . . Developing semiautomatic tools for learning ontologies and extracting metadata is a key research area . . . .Having to move out of a user’s typical working environment to ‘do knowledge management’ will act as a disincentive, whether the user is creating or retrieving knowledge.

Of course, even assuming that ontologies are created and semantics and metadata are added to content, there still remains the nasty problems of resolving heterogeneities (semantic mediation) and efficiently storing and retrieving the metadata and semantic relationships.

Putting all of this process in place requires the infrastructure in the form of tools and automation and proper incentives and rewards for users and suppliers to conform to it.

Areas Requiring Tools and Automation

In his paper, Warren repeatedly points to the need for “semi-automatic” methods to make the semantic Web a reality. He makes fully a dozen such references, in addition to multiple references to the need for “reasoning algorithms.” In any case, here are some of the areas noted by Warren needing “semi-automatic” methods:

• Assign authoritativeness
• Learn ontologies
• Infer better search requests
• Mediate ontologies (semantic resolution)
• Support visualization
• Assign collaborations
• Infer relationships
• Extract entities
• Create ontologies
• Maintain and evolve ontologies
• Create taxonomies
• Infer trust
• Analyze links
• etc.

In a different vein, SemWebCentral lists these clusters of semantic Web-related tasks, each of which also requires tools:[2]

• Create an ontology — use a text or graphical ontology editor to create the ontology, which is then validated. The resulting ontology can then be viewed with a browser before being published
• Disambiguate data – generate a mapping between multiple ontologies to identify where classes and properties are the same
• Expose a relational database as OWL — an editor is first used to create the ontologies that represent the database schema, then the ontologies are validated, translated to OWL and then the generated OWL is validated
• Intelligently query distributed data – repository and again able to be queried
• Manually create data from an ontology — a user would use an editor to create new OWL data based on existing ontologies, which is then validated and browsable
• Programmatically interact with OWL content — custom programs can view, create, and modify OWL content with an API
• Query non-OWL data — via an annotation tool, create OWL metadata from non-OWL content
• Visualize semantic data — view semantic data in a custom visualizer.

With some ontologies approaching tens to hundreds of thousands to millions of triples, viewing, annotating and reconciling at scale can be daunting tasks, the efforts behind which would never be taken without useful tools and automation.

A Workflow Perspective Helps Frame the Challenge

A 2005 paper by Izza, Vincent and Burlat (among many other excellent ones) at the first International Conference on Interoperability of Enterprise Software and Applications (INTEROP-ESA) provides a very readable overview on the role of semantics and ontologies in enterprise integration.[3] Besides proposing a fairly compelling unified framework, the authors also present a useful workflow perspective emphasizing Web services (WS), also applicable to semantics in general, that helps frame this challenge:

Generic Semantic Integration Workflow (adapted from [3])

For existing data and documents, the workflow begins with information extraction or annotation of semantics and metadata (#1) in accordance with a reference ontology. Newly found information via harvesting must also be integrated; however, external information or services may come bearing their own ontologies, in which case some form of semantic mediation is required.

Of course, this is a generic workflow, and depending on the interoperation task, different flows and steps may be required. Indeed, the overall workflow can vary by perspective and researcher, with semantic resolution workflow modeling a prime area of current investigations. (As one alternative among scores, see for example Cardoso and Sheth.[4])

Matching and Mapping Semantic Heterogeneities

Semantic mediation is a process of matching schemas and mapping attributes and values, often with intermediate transformations (such as unit or language conversions) also required. The general problem of schema integration is not new, with one prior reference going back as early as 1986. [5] According to Alon Halevy:[6]

As would be expected, people have tried building semi-automated schema-matching systems by employing a variety of heuristics. The process of reconciling semantic heterogeneity typically involves two steps. In the first, called schema matching, we find correspondences between pairs (or larger sets) of elements of the two schemas that refer to the same concepts or objects in the real world. In the second step, we build on these correspondences to create the actual schema mapping expressions.

The issues of matching and mapping have been addressed in many tools, notably commercial ones from MetaMatrix,[7] and open source and academic projects such as Piazza, [8] SIMILE, [9] and the WSMX (Web service modeling execution environment) protocol from DERI. [10] [11] A superb description of the challenges in reconciling the vocabularies of different data sources is also found in the thesis by Dr. AnHai Doan, which won the 2003 ACM’s Prestigious Doctoral Dissertation Award.[12]

What all of these efforts has found is the inability to completely automate the mediation process. The current state-of-the-art is to reconcile what is largely unambiguous automatically, and then prompt analysts or subject matter experts to decide the questionable matches. These are known as “semi-automated” systems and the user interface and data presentation and workflow become as important as the underlying matching and mapping algorithms. According to the WSMX project, there is always a trade-off between how accurate these mappings are and the degree of automation that can be offered.

Also a Need for Efficient Semantic Data Stores

Once all of these reconciliations take place there is the (often undiscussed) need to index, store and retrieve these semantics and their relationships at scale, particularly for enterprise deployments. This is a topic I have addressed many times from the standpoint of scalability, more scalability, and comparisons of database and relational technologies, but it is also not a new topic in the general community.

As Stonebraker and Hellerstein note in their retrospective covering 35 years of development in databases,[13] some of the first post-relational data models were typically called semantic data models, including those of Smith and Smith in 1977[14] and Hammer and McLeod in 1981.[15] Perhaps what is different now is our ability to address some of the fundamental issues.

At any rate, this subsection is included here because of the hidden importance of database foundations. It is therefore a topic often addressed in this series.

A Partial Listing of Semantic Web Tools

In all of these areas, there is a growing, but still spotty, set of tools for conducting these semantic tasks. SemWebCentral, the open source tools resource center, for example, lists many tools and whether they interact or not with one another (the general answer is often No).[16] Protégé also has a fairly long list of plugins, but not unfortunately well organized. [17]

In the table below, I begin to compile a partial listing of semantic Web tools, with more than 50 listed. Though a few are commercial, most are open source. Also, for the open source tools, only the most prominent ones are listed (Sourceforge, for example, has about 200 projects listed with some relation to the semantic Web though most of minor or not yet in alpha release).

Tools Still Crude, Integration Not Compelling

Individually, there are some impressive and capable tools on this list. Generally, however, the interfaces are not intuitive, integration between tools is lacking, and why and how standard analysts should embrace them is lacking. In the semantic Web, we have yet to see an application of the magnitude of the first Mosaic browser that made HTML and the World Wide Web compelling.

It is perhaps likely that a similar “killer app” may not be forthcoming for the semantic Web. But it is important to remember just how entwined tools are to accelerating acceptance and growth of new standards and protocols.

This Friday brown bag leftover was first placed into the AI3 refrigerator about four years ago on June 12, 2006. It was the follow-on to last week’s Brown Bag Lunch posting. It is also the first attempt I made at assembling semantic Web- and -related tools, which has now grown into the 800+ Sweet Tools listing. No changes have been made to the original posting.

Earlier postings in this recent series traced the progress in climbing the data federation pyramid to today’s current emphasis on the semantic Web. Partially this series is aimed at disabusing the notion that data extensibility can arise simply by using the XML (eXtensible Markup Language) data representation protocol. As Stonebraker and Hellerstein correctly observe:

XML is sometimes marketed as the solution to the semantic heterogeneity problem . . . . Nothing could be further from the truth. Just because two people tag a data element as a salary does not mean that the two data elements are comparable. One could be salary after taxes in French francs including a lunch allowance, while the other could be salary before taxes in US dollars. Furthermore, if you call them “rubber gloves” and I call them “latex hand protectors”, then XML will be useless in deciding that they are the same concept. Hence, the role of XML will be limited to providing the vocabulary in which common schemas can be constructed.[1]

This series also covers the ontologies and the OWL language (written in XML) that now give us the means to understand and process these different domains and “world views” by machine. According to Natalya Noy, one of the principal researchers behind the Protégé development environment for ontologies and knowledge-based systems:

How are ontologies and the Semantic Web different from other forms of structured and semi-structured data, from database schemas to XML? Perhaps one of the main differences lies in their explicit formalization. If we make more of our assumptions explicit and able to be processed by machines, automatically or semi-automatically integrating the data will be easier. Here is another way to look at this: ontology languages have formal semantics, which makes building software agents that process them much easier, in the sense that their behavior is much more predictable (assuming they follow the specified explicit semantics–but at least there is something to follow). [2]

Again, however, simply because OWL (or similar) languages now give us the means to represent an ontology, we still have the vexing challenge of how to resolve the differences between different “world views,” even within the same domain. According to Alon Halevy:

When independent parties develop database schemas for the same domain, they will almost always be quite different from each other. These differences are referred to as semantic heterogeneity, which also appears in the presence of multiple XML documents, Web services, and ontologies–or more broadly, whenever there is more than one way to structure a body of data. The presence of semi-structured data exacerbates semantic heterogeneity, because semi-structured schemas are much more flexible to start with. For multiple data systems to cooperate with each other, they must understand each other’s schemas. Without such understanding, the multitude of data sources amounts to a digital version of the Tower of Babel. [3]

In the sections below, we describe the sources for how this heterogeneity arises and classify the many different types of heterogeneity. I then describe some broad approaches to overcoming these heterogeneities, though a subsequent post looks at that topic in more detail.

Causes and Sources of Semantic Heterogeneity

There are many potential circumstances where semantic heterogeneity may arise (partially from Halevy [3]):

• Enterprise information integration
• Querying and indexing the deep Web (which is a classic data federation problem in that there are literally tens to hundreds of thousands of separate Web databases) [4]
• Merchant catalog mapping
• Schema v. data heterogeneity
• Schema heterogeneity and semi-structured data.

Naturally, there will always be differences in how differing authors or sponsors create their own particular “world view,” which, if transmitted in XML or expressed through an ontology language such as OWL may also result in differences based on expression or syntax. Indeed, the ease of conveying these schema as semi-structured XML, RDF or OWL is in and of itself a source of potential expression heterogeneities. There are also other sources in simple schema use and versioning that can create mismatches [3]. Thus, possible drivers in semantic mismatches can occur from world view, perspective, syntax, structure and versioning and timing:

• One schema may express a similar “world view” with different syntax, grammar or structure
• One schema may be a new version of the other
• Two or more schemas may be evolutions of the same original schema
• There may be many sources modeling the same aspects of the underlying domain (”horizontal resolution” such as for competing trade associations or standards bodies), or
• There may be many sources that cover different domains but overlap at the seams (”vertical resolution” such as between pharmaceuticals and basic medicine).

Regardless, the needs for semantic mediation are manifest, as are the ways in which semantic heterogeneities may arise.

Classification of Semantic Heterogeneities

The first known classification scheme applied to data semantics that I am aware of is from William Kent nearly 20 years ago.[5] (If you know of earlier ones, please send me a note.) Kent’s approach dealt more with structural mapping issues (see below) than differences in meaning, which he pointed to data dictionaries as potentially solving.

The most comprehensive schema I have yet encountered is from Pluempitiwiriyawej and Hammer, “A Classification Scheme for Semantic and Schematic Heterogeneities in XML Data Sources.” [6] They classify heterogeneities into three broad classes:

• Structural conflicts arise when the schema of the sources representing related or overlapping data exhibit discrepancies. Structural conflicts can be detected when comparing the underlying DTDs. The class of structural conflicts includes generalization conflicts, aggregation conflicts, internal path discrepancy, missing items, element ordering, constraint and type mismatch, and naming conflicts between the element types and attribute names.
• Domain conflicts arise when the semantic of the data sources that will be integrated exhibit discrepancies. Domain conflicts can be detected by looking at the information contained in the DTDs and using knowledge about the underlying data domains. The class of domain conflicts includes schematic discrepancy, scale or unit, precision, and data representation conflicts.
• Data conflicts refer to discrepancies among similar or related data values across multiple sources. Data conflicts can only be detected by comparing the underlying DOCs. The class of data conflicts includes ID-value, missing data, incorrect spelling, and naming conflicts between the element contents and the attribute values.

Moreover, mismatches or conflicts can occur between set elements (a “population” mismatch) or attributes (a “description” mismatch).

The table below builds on Pluempitiwiriyawej and Hammer’s schema by adding the fourth major explicit category of language, leading to about 40 distinct potential sources of semantic heterogeneities:

Most of these line items are self-explanatory, but a few may not be:

• Homonyms refer to the same name referring to more than one concept, such as Name referring to a person v. Name referring to a book
• A generalization/specialization mismatch can occur when single items in one schema are related to multiple items in another schema, or vice versa. For example, one schema may refer to “phone” but the other schema has multiple elements such as “home phone,” “work phone” and “cell phone”
• Intra-aggregation mismatches come when the same population is divided differently (Census v. Federal regions for states, or full person names v. first-middle-last, for examples) by schema, whereas inter-aggregation mismatches can come from sums or counts as added values
• Internal path discrepancies can arise from different source-target retrieval paths in two different schema (for example, hierarchical structures where the elements are different levels of remove)
• The four sub-types of schematic discrepancy refer to where attribute and element names may be interchanged between schema
• Under languages, encoding mismatches can occur when either the import or export of data to XML assumes the wrong encoding type. While XML is based on Unicode, it is important that source retrievals and issued queries be in the proper encoding of the source. For Web retrievals this is very important, because only about 4% of all documents are in Unicode, and earlier BrightPlanet provided estimates there may be on the order of 25,000 language-encoding pairs presently on the Internet
• Even should the correct encoding be detected, there are significant differences in different language sources in parsing (white space, for example), syntax and semantics that can also lead to many error types.

It should be noted that a different take on classifying semantics and integration approaches is taken by Sheth et al. [7] Under their concept, they split semantics into three forms: implicit, formal and powerful. Implicit semantics are what is either largely present or can easily be extracted; formal languages, though relatively scarce, occur in the form of ontologies or other descriptive logics; and powerful (soft) semantics are fuzzy and not limited to rigid set-based assignments. Sheth et al.’s main point is that first-order logic (FOL) or descriptive logic is inadequate alone to properly capture the needed semantics.

From my viewpoint, Pluempitiwiriyawej and Hammer’s [6] classification better lends itself to pragmatic tools and approaches, though the Sheth et al. approach also helps indicate what can be processed in situ from input data v. inferred or probabalistic matches.

Importance of Reference Standards

An attractive and compelling vision  — perhaps even a likely one  — is that standard reference ontologies become increasingly prevalent as time moves on and semantic mediation is seen as more of a mainstream problem. Certainly, a start on this has been seen with the use of the Dublin Core metadata initiative, and increasingly other associations, organizations, and major buyers are busy developing “standardized” or reference ontologies.[8] Indeed, there are now more than 10,000 ontologies available on the Web.[9] Insofar as these gain acceptance, semantic mediation can become an effort mostly at the periphery and not the core.

But, such is not the case today. Standards only have limited success and in targeted domains where incentives are strong. That acceptance and benefit threshold has yet to be reached on the Web. Until such time, a multiplicity of automated methods, semi-automated methods and gazetteers will all be required to help resolve these potential heterogeneities.

This Friday brown bag leftover was first placed into the AI3 refrigerator about four years ago on June 6, 2006. No changes have been made to the original posting. Current approaches to dealing with these heterogeneities would be to use “bridging” ontologies that map the mismatches.

Today, in the advanced knowledge economy of the United States, the information contained within documents represents about a third of total gross domestic product, or an amount of about $3.3 trillion annually.

Yet our understanding of the value of documents and the means to manage them is abysmal. These failures impact enterprises of all sizes from the standpoints of revenues, profitability and reputation. Continued national productivity growth — and thus the wealth of all citizens — depends critically on understanding and managing these document values.

As this white paper describes, the lack of a compelling and demonstrable common understanding of the importance of documents is in itself a major factor limiting available productivity benefits. There is an old Chinese saying that roughly translated is “what cannot be measured, cannot be improved.” Many corporate officers may believe this to be the case for document creation and productivity, but, as this paper shows, in fact many of these document issues can be measured.

This Friday brown bag leftover was first placed into the AI3 refrigerator on July 20, 2005. No changes have been made to the original posting.

I’d like to thank David Siegel for recently highlighting this post from 5 years ago with nice kudos on his PowerOfPull blog. That reference is what caused me to dust off the cobwebs from this older piece.

To wit, some 25% of all of the annual trillions of dollar spent on document creation costs lend themselves to actionable improvements:

Table 1. Mid-range Estimates for the Annual Value of Documents, U.S. Firms, 2002[1]

The total benefit from improved document access and use to the U.S economy is on the order of $800 billion annually, or about 8% of GDP. For the 1,000 largest U.S. firms, benefits from these improvements can approach nearly $250 million annually per firm. About three-quarters of these benefits arise from not re-creating the intellectual capital already invested in prior document creation. About one-quarter of the benefits are due to reduced regulatory non-compliance or paperwork, or better competitiveness in obtaining solicited grants and contracts.

Indeed, even these figures likely severely underestimate the benefits to enterprises from an improved leverage of document assets. It has always been the case that the best and most successful companies have been able to make better advantage of their intellectual assets than their competitors. The competitiveness advantage from better document access and use alone may exceed the huge benefits in the table above.

Documents — that is, unstructured and semi-structured data — are now at the point where structured data was at 15 years ago. At that time, companies realized that consolidating information from multiple numeric databases would be a key source of competitive advantage. That realization led to the development and growth of the data warehousing or business intelligence markets, now representing about $3.9 billion in annual software sales.

Search and enterprise content management software today only represents a fraction of that amount — perhaps on the order of $500 million annually. But given that intellectual content in documents represents three to four times the amount in numeric structured data, it is clear that document software capabilities are not being well utilized, reaching only a small fraction of their market potential.

The estimates provided in this white paper are drawn from numerous sources and are extremely fragmented, perhaps even inconsistent. One hope in preparing this document was to stimulate more research attention and data gathering around the critical issues of document value to the enterprise and the economy at large.

EXECUTIVE SUMMARY

I. INTRODUCTION

Documents: The Drivers of a Knowledge Economy

Documents: The Linchpin of Corporate Intellectual Assets

Documents: Unknown Value, Huge Implications

Documents: The Next Generation of Data Warehousing?

Connecting the Dots: A Pointillistic Approach

II. INTERNAL DOCUMENTS

Number of ‘Valuable’ Documents Produced per Firm

Total Annual U.S. ‘Costs’ to Create Documents

‘Cost’ of Creating a ‘Typical’ Document

‘Cost’ of a Missed or Overlooked Document

Other Document Total ‘Cost’ Factors and Summary

Archival Lifetime of ‘Valuable’ Documents

III. WEB DOCUMENTS AND SEARCH

Estimate of Time and Effort Devoted to Document Search

Effect of Non-persistent Search Efforts

‘Cost’ of Creating and Maintaining a Document Category Portal

‘Cost’ of Inaccessible or Hidden Intranet Sites

IV. OPPORTUNITIES AND THREATS

‘Costs’ and Opportunity Costs of Winning Proposals

‘Costs’ of Regulation and Regulatory Non-compliance

‘Cost’ of an Unauthorized Posted Document

V. CONCLUSIONS

I. INTRODUCTION

How many documents does your organization create each year? What effort does this represent in terms of total staffing costs? What does it cost to create a ‘typical’ document? Of documents created, how much of the value in them is readily sharable throughout your organization? How long do you need to keep valuable documents and how can you access them? How much existing document content is re-created simply because prior work cannot be found? When prior information is missed, what do these prior investments in documents represent in terms of loss of market share, revenue or reputation? Indeed, what does the term, “document” represent in your organization’s context?

If you have difficulty answering these questions, you are not alone. Depending on the survey, from 90% to 97% of enterprises cannot answer these questions — in whole or in part. The purpose of this white paper is to provide the first comprehensive assessment ever of these document values.

Enterprises and the analyst community have historically overlooked the impact of document creation as opposed to document handling. Document creation is about 2-3 times more important — from an embedded cost standpoint — than document handling. Second, all aspects of document creation, and later access and use, assume a much greater role in the overall economics of enterprises than have been realized previously.

Documents: The Drivers of a Knowledge Economy

Put your index finger one inch from your nose. That is how close — and unfocused — document importance is to an organization. Documents are the salient reality of a knowledge economy, but like your finger, documents are often too close, ubiquitous and commonplace to appreciate.

How do your employees earn their livings? Writing proposals? Marketing or selling? Evaluating competitors or opportunities? Persuading? Analyzing? Communicating? Teaching? Of course, in some sectors, many make their living from growing things or making things. These are essential jobs — indeed, until the last few decades were the predominant drivers of economies — but are now being supplanted in advanced economies by knowledge work. Perhaps up to 35% of all company employees in the U.S. can be classified as knowledge workers.

And knowledge work means documents. The fact is that knowledge is produced and communicated through the written word. When we search, when we write, when we persuade, we may often do so verbally but make it persistent through the written word.

Documents: The Linchpin of Corporate Intellectual Assets

IBM estimates that corporate data doubles every six to eight months, 85% of which are documents.[2] At least 10% of an enterprise’s information changes on a monthly basis.[3] Year-on-year office document growth rates are on the order of 22%.[4] As later analysis indicates, there are perhaps on the order of 10 billion documents created annually in the U.S with a mid-range “asset” value of $3.3 trillion per year. Documents are a huge contributor to the United States’ gross domestic product of $10.5 trillion (2002).

• According to a Coopers & Lybrand study in 1993:[5]
• Ninety percent of corporate memory exists on paper
• Ninety percent of the papers handled each day are merely shuffled
• Professionals spend 5-15 percent of their time reading information, but up to 50 percent looking for it
• On average, 19 copies are made of each paper document.

A Xerox Corporation study commissioned in 2003 and conducted by IDC surveyed 1000 of the largest European companies and had similar findings:[6],[7]

• On average 45% of an executive’s time was spent dealing with documents
• 82% believe that documents were crucial to the successful operation of their organizations
• A further 70% claimed that poor document processes could impact the operational agility of their organizations
• While 83%, 78% and 76% consider faxes, email and electronic files as documents, respectively, only 48% and 46% categorize web pages and multimedia content as such.

Documents: Unknown Value, Huge Implications

But, if defining what constitutes a document is hard, identifying the costs associated with all the document activities is almost impossible for many organizations. Ninety to 97 percent of the corporate respondents to the Coopers & Lybrand and Xerox studies, respectively, could not estimate how much they spent on producing documents each year. Almost three quarters of them admit that the information is unavailable or unknown to them.

An A.T. Kearney study sponsored by Adobe, EDS, Hewlett-Packard, Mayfield and Nokia, published in 2001, estimated that workforce inefficiencies related to content publishing cost organizations globally about $750 billion. The study further estimated that knowledge workers waste between 15% to 25% of their time in non-productive document activities.[8]

Figure 1. The Situation of Poor Enterprise Document Use Leads to Real Implications

But the situation is much broader and results in part from the inability to quantify the importance of both internal and external document assets to all aspects of the enterprise’s bottom line. For examples drawn from the main body of this white paper, early adopters of enterprise content software typically capture less than 1% of valuable internal documents available; large enterprises are witnessing the proliferation of internal and external Web sites, sometimes exceeding thousands; use of external content is presently limited to Internet search engines, producing non-persistent results and no capture of the investment in discovery or results; and “deep” content in searchable databases, which is common to large organizations and represents 90% of external Internet content, is completely untapped.

A USC study reported that typically only 32% of employees in knowledge organizations have access to good information about technical developments relevant to their work, and 79% claim they have inadequate information about what their competitors are doing.[9]

The enterprise content integration software market is fragmented and confused, with only a few established companies providing partial solutions. Content integration is still a small market with annual revenues of less than $50 million worldwide.[10] Vendor offerings fail to satisfy customer needs because of a lack of functionality and a lack of scalability to enterprise volumes. Sales in the market remain distinctly lower than those projected by industry analysts, even as the magnitude of “information overload” continues to grow at a dramatic rate.

Documents: The Next Generation of Data Warehousing?

Documents — that is, unstructured and semi-structured data — are now at the point where structured data was at 15 years ago. At that time, companies realized that consolidating information from multiple numeric databases would be a key source of competitive advantage. That realization led to the development and growth of the data warehousing or business intelligence markets, now representing about $3.9 billion in annual software sales.[11]

Certain categories of businesses have been leaders in content integration, especially those that have recently had mergers and acquisitions activity, those that need to integrate business applications with content, and those for which the reuse of marketing assets across the organization is critical.¹⁰

Stonebraker and Hellerstein have provided an insightful roadmap for how enterprise data integration or “federation” has trended over time: Data warehousing → Enterprise application integration → Enterprise content integration → Enterprise information integration.[12] There are two threads to this trend. First, there has been a growing recognition of the importance of document (unstructured) content to contribute to actionable information. Second, increasingly unified and integrated means are being applied to all data sources to allow single-access retrievals.

Connecting the Dots: A Pointillistic Approach

The state of information regarding the value and cost of documents is extremely poor. Lack of defensible and vetted estimates for this information undercuts the ability to properly estimate the intellectual assets tied up in documents or the impacts of overlooked or misused documents.

Only three large document studies — the Coopers & Lybrand, Xerox and A.T. Kearney studies noted above — have been conducted in the past ten years regarding the use and importance of documents within enterprises, and then solely from the standpoint of executive perceptions.

The quantified picture presented in this white paper regarding the costs and benefits of document creation, access and use is a paint-by-the-numbers assemblage of disparate data. The paper draws upon about 80 different data sources, many fragmented. The analysis approach by necessity has needed to conjoin assumptions and data from many diverse sources.

This approach leads to both uncertainty regarding “true” values and likely inaccuracies or mis-estimates in some areas. To make the assessment as consistent as possible, a base year of 2002 was used, the common year reference for most of the available data sources. To bracket uncertainties, most estimates are provided in low, medium and high estimates.

Thus, this study should be viewed as preliminary, but strongly indicative of the value of documents. Further research and data collection will surely refine these estimates. Clearly, though, by any measure, the value of documents to the enterprise is significant and huge, and should not continue to be overlooked.

II. INTERNAL DOCUMENTS

Though valuable content resides everywhere, the first challenge to enterprises is getting a handle on their own internal document content.

Number of ‘Valuable’ Documents Produced per Firm

A recent UC Berkeley study on “How Much Information?” estimated that more than 4 billion pages of internal office documents with archival value are generated annually in the U.S. (Note: this is not the amount created, only those documents deemed worthy of retaining for more than one year).

Table 2. Document Projections for U.S. Firms by Size, 2002 Basis

Sources: UC Berkeley[13], U.S. Commerce Department[14], U.S. Bureau of Labor Statistics[15], U.S. Census Bureau[16]

Table 2 and Table 3 attempt to summarize the scale of this challenge for U.S. firms (for internal enterprise documents only). (See[17] for a description of methodology regarding document scales, note[18] for estimating the numbers of enterprise knowledge workers, and note[19] for estimating content workers. A rough multiplier of 3x to 4x can be applied to extrapolate globally.[20]) Breakouts are provided by size of firm; these include estimates for the number of knowledge and content workers within U.S. firms.

Table 3. Total Annual Document Projections for U.S. Firms, 2002 Basis

Table 4 takes this information and breaks out distribution of document production for a ‘typical’ knowledge worker according to major document types. The data from this table is based on analysis of dozens of BrightPlanet customers averaged across about 10 million documents in various repositories.

Table 4. Document Production for a ‘Typical’ Knowledge Worker

Note that word processed documents account for about 50% of typical production and storage demands. However, also note that documents of the highest archival value, as converted to PDFs for sharing and deployment, also represent about a third to two-fifths of stored documents.

Total Annual U.S. ‘Costs’ to Create Documents

Based on the information from Table 2 to Table 4 above, all updated to a common year 2002 basis, we can now estimate the total annual costs in the U.S. for creating all internal enterprise documents. The analysis is based on the UC Berkeley information and the Coopers & Lybrand studies. The “bottom up” case is based on the number of annual U.S. documents estimated based on Table 2. These results are shown in the table below:

Table 5. Annual U.S. Office Document Cost Estimates[21]

The average numbers above represent the average of the unique values in each column. The Table 5 analysis suggests there may be on the order of 10 billion documents created annually in the U.S with a total “asset” value on the order of $3.3 trillion per year.

‘Cost’ of Creating a ‘Typical’ Document

Based on the averages in the table above, a ‘typical’ document may cost on the order of $380 each to create.[22] Of course, a “document” can vary widely in size, complexity and time to create, and therefore its individual cost and value will vary widely. An invoice generated from an automated accounting system could be a single page and produced automatically in the thousands; proposals for very large contracts can take tens of thousands to millions of dollars to create. For examples, here are some other ‘typical’ costs for a variety of documents:

Table 6. ‘Typical’ per Document Creation Costs

Depending on document mix and activities, individual enterprises may want to vary the average document creation costs used in their cost-benefit estimates.

‘Cost’ of a Missed or Overlooked Document

The Coopers & Lybrand study suggests that 7.5 percent of all documents are lost forever, and that it costs $120 in labor ($150 updated to 2002) to find a misfiled document;[26] other studies suggest that 5% to 6% of documents are routinely misplaced or misfiled.

In fact, the extent of this problem is unknown and is affirmed by the Xerox results:[27]

• Almost three quarters of corporate respondents admit that the information is unavailable or unknown to them
• 95% of the companies are not able to estimate the cost of wasted or unused documents
• On average 19% of printed documents were wasted.

Other Document Total ‘Cost’ Factors and Summary

Five independent studies suggest that, on average, organizations spend from 5% to 15% of total company revenue on handling documents.^{27,[28],[29],[30],[31]} These seemingly innocuous percentages can translate into huge bottom-line impacts for U.S. enterprises. For example, the total GDP of the United States was on the order of $10.5 trillion at the end of 2002.[32] Translating this value into the results of Table 5 and the information in previous sections indicates the importance of document creation and handling for U.S enterprises:

Table 7. Range Estimates for Total U.S. Document Burdens in Enterprises, 2002[33]

A few observations relate to this table. First, enterprises and the analyst community have greatly overlooked the impact of document creation as opposed to document handling. Document creation is about 2-3 times more important  – from an embedded cost standpoint  – than document handling. Second, all aspects of document creation assume a much greater role in the overall economics of enterprises than has been realized previously.

The fact that documents have received so little management attention, awareness, measurement and direct attention to improve performance is shocking.

Archival Lifetime of ‘Valuable’ Documents

The ‘low’ and ‘high’ estimates for documents in Table 2 and Table 3 assume that 2% and 5%, respectively, of internal documents have archival value. Were these percentages to be higher, the volume of documents requiring integration and access would likewise increase. The 2% value is derived from the UC Berkeley study,[34] which also refers to an unpublished European study that places archival amounts at 10%. Unfortunately, there is little empirical information to support the degree to which documents deserve to be kept for archival purposes.

Assuming that documents may retain value for three to five years, the largest firms perhaps have as many as 4 million internal documents on average with enterprise-wide value. Firms with fewer employees generally have lower document counts. Archival percentages, however, are a tricky matter, since apparently 85% of all archived documents are accessed.[35]

III. WEB DOCUMENTS AND SEARCH

Various estimates by Cowles/Simba,[36] Veronis, Suhler & Associates,[37] and Outsell[38] place the current market for on line business information in the $30 billion to $140 billion range, with significant projected growth. Outsell also indicates that marketing, sales, and product development professionals rely most heavily on information from the Internet for their daily decision making, based on a comparative study of Fortune 500 business professionals’ use of the open Web and fee-based desktop information content services.[39] Clearly, relevant and targeted content, much of which resides on line, has extreme value to enterprises.

UC Berkeley estimates that about 500 petabytes of new information was published on the Web in 2002,³⁴ based on original analysis conducted by BrightPlanet.[40] The compound growth rate in Web documents has been on the order of more than 200% annually.[41] Estimates for deep Web content range from about 6-8 times larger [42] to 500 times larger40 than standard “surface web” content. The size of Internet content is overwhelming, of highly variable quality, growing at a rapid pace, and with much of its content ephemeral.

Estimate of Time and Effort Devoted to Document Search

According to a recent study by iProspect, about 56 percent of users use search engines every day, based on a population of which more than 70 percent use the Internet more than 10 hours per week. Professionals abandon a current search 38% of the time after inspecting only one results page (the listing of document result URLs), and overall 82% of users attempt another search if relevant results are not found within the first three results pages. Just 13 percent of users said that they use different search engines for different types of searches.[43] Only 7.5 percent of Internet users said they refined their search with additional keywords in cases where they were unable to achieve satisfactory results.[44]

The average knowledge worker spends 2.3 hrs per day  – or about 25% of work time  – searching for critical job information.[45] IDC estimates that enterprises employing 1,000 knowledge workers waste well over $6 million per year each in searching for information that does not exist, failing to find information that does, or recreating information that could have been found but was not.[46] As that report stated, “It is simply impossible to create knowledge from information that cannot be found or retrieved.”

Vendors and customers often use time savings by knowledge workers as a key rationale for justifying a document or content initiative. This comes about because many studies over the years have noted that white collar employees spend a consistent 20% to 25% of their time seeking information; the premise is that more effective search will save time and drop these percentages. As a sample calculation, each 1% reduction in time devoted to search produces:

$50,000 (base salary) * 1.8 (burden rate) * 1.0% = $900/ employee

The stable percentage effort devoted to search over time suggests it is the “satisficing” allocation. (In other words, knowledge workers are willing to devote a quarter of their time to finding relevant information.) Thus, while better tools to aid better discovery may lead to finding better information and making better decisions more productively  – a far more important justification in itself  – there may not result a strict time or labor savings from more efficient search.[47]

Effect of Non-persistent Search Efforts

The percentage of Web page visits that are re-visits is estimated at between 58%[48] and 80%.[49] While many of these re-visitations occur shortly after the first visit (e.g., during the same session using the back button), a significant number occur after a considerable amount of time has elapsed. Thus, it is not surprising that a survey of problems using the Web found “Not being able to find a page I know is out there,” and “Not being able to return to a page I once visited,” accounted for 17% of the problems reported, and that the most common problem using bookmarks was, “Changed content.”[50] Depending on the content type, users use either “direct” or “indirect” approaches to re-find previously discovered information:

Table 8. General Approaches to Re-finding Previously Discovered Information [51]

Direct approaches require remembering or specifically noting the specific location of the information. Direct approaches include: direct entry; emailing to self; emailing to others; printing out; saving as file; pasting the URL into a document; and posting to a personal Web site.

Indirect approaches include: searching; looking through bookmarks; and recalling from a history file. All of these indirect approaches are supported by modern browsers. Note that re-finding Web pages or documents relies heavily on having a record of a previously visited URL.

As a University of Washington study supported by Microsoft discovered, all of the specific direct and indirect techniques applied to these re-discovery approaches have significant drawbacks in terms of desired functions for the recall process: [52]

Table 9. Strengths and Weakness of Existing Techniques to Re-use Web Information

The general observation is that no present technique is able alone to keep search persistent, current or maintain context. These combined inadequacies mean that previously found information is not easily found again, or re-discovered, as the following table shows:

Table 10. Success in Finding Important Earlier Found Web Information [53]

This table has a number of important observations. First, some 37% of previously found information disappears from the Web, consistent with other findings that estimate about 40% of all Web content disappears annually, some of which has historical or archival value.[54]

Second, and most importantly, nearly 70% of previously found valuable information cannot be rediscovered again. More than half of this problem is because the information is no longer available on the Web, but other reasons relate to the inadequacies of recall techniques for finding previously discovered information.

These observations can translate into some relatively huge costs on a per employee and per enterprise basis, as the table below shows:

Table 11. ‘Cost’ of Not Readily Re-finding Valuable Web Information

This analysis assumes that some previously found information of value is again re-found (60%), but some is also not re-found and must be re-created (40%).[55] The ‘large’ enterprise is identical to the definition in Table 2 (which is also nearly equivalent to a Fortune 1000 company).[56]

The analysis indicates that poor methods to recall previously found and valuable Web documents may cost $1,600 per knowledge worker per year. This translates into nearly a $10 million productivity loss for the largest enterprises, or nearly $33 billion across all U.S. industries.

In relation to the total document costs noted in Table 7 above, these may seem to be comparatively small numbers. However, when viewed in the context of unproductive standard Web search, they indicate important failings in the ability to recall previously found valuable results from searches and their attendant productivity losses.

‘Cost’ of Creating and Maintaining a Document Category Portal

Users, administrators and industry analysts alike recognize the importance of placing content into logical, intuitive and hierarchically organized categories. About 60% of knowledge workers note that search is a difficult process, made all the more difficult without a logical organization to content.[57] While technical distinctions exist, these logical structures organized into a hierarchical presentation are most often referred to as “taxonomies,” though other terms such as ontology, subject directory, subject tree, directory structure or classification schema may be used.

Delphi Group’s research with corporate Web sites points to the lack of organized information as the number one problem in the opinion of business professionals. More than three-quarters of the surveyed corporations indicated that a taxonomy or classification system for documents is imperative or somewhat important to their business strategy; more than one-third of firms that classify documents still use manual techniques.⁵⁷ Hierarchical arrangements of categorized subjects trigger associations and relationships that are not obvious when simply searching keywords. Other advantages cited for the taxonomic presentation of documents are the greater likelihood of discovery, ease-of-use, overcoming the difficulty of formulating effective search queries, being able to search only within related documents, discovery of relationships among similar terminology and concepts, and user satisfaction.[58],[59]

From the user standpoint, knowledge workers want to impose taxonomic order on document chaos, but only if the taxonomy models their domain accurately. They also want software to assist with categorizing, as long as it respects the taxonomy they created. Finally, the results of these category placements should be presented via a portal. Thus, as the common concern across all requirements, the taxonomy takes on tremendous importance for an application’s success.[60]

Figure 2. Typical Large Firm Documents, Thousands

Enterprises that have adopted directory structures for content management are not yet achieving enterprise-wide relevance, presenting on average 1% of all relevant documents in an organized portal view. These limitations appear to be driven by weaknesses in the technology and high costs associated with conventional approaches:

• Comprehensiveness and Scale – according to a market report published by Plumtree in 2003, the average document portal contains about 37,000 documents.[61] This was an increase from a 2002 Plumtree survey that indicated average document counts of 18,000.[62] However, about 60% of respondents to a Delphi Group survey said they had more than 50,000 internal documents in their portal environment (generally the department level), ³ and as Table 2 indicates above, most of the largest firms likely have millions or more internal documents deserving of common access and archiving.
• The left-hand bar in Figure 2 indicates current averages for documents in existing content portals. The right-hand (yellow and orange) bar indicates potential based on high and low estimates. The ‘Archive’ case (middle bar) show the same values as provided in Table 2, and represent a conservative view of “archival-likely” documents. The right bar is a more representative view of actual current internal content that enterprises may want to make available to their employees.[63] Two observations have merit: 1) under current practice, enterprises are at most making 10% of their useful documents available, and more likely slightly over 1%; 2) the documents that are being made available are solely internal, and neglect potentially important external sources that would increase document counts considerably.
• Implementation Times – though average time to stand-up a new content installation is about 6 months, there is also a 22% risk that deployment times exceeds that and an 8% risk it takes longer than one year. Furthermore, internal staff necessary for initial stand-up average nearly 14 people (6 of whom are strictly devoted to content development), with the potential for much larger head counts[64]
• Ongoing Maintenance and Staffing Costs – ongoing maintenance and staffing costs typically exceed the initial deployment effort. This trend is perhaps not surprising in that once a valuable content portal has been created there will be demands to expand its scope and coverage. Based on these various factors, Table 12 summarizes set-up, ongoing maintenance and key metrics for today’s conventional approaches versus what BrightPlanet can do (the BrightPlanet document count is based on a ‘typical’ installation; there are no practical scale limits)

Table 12. Staff, Time and per Document Costs for Categorized Document Portals

• The content staff level estimates in the table are consistent with anecdotal information and with a survey of 40 installations that found there were on average 14 content development staff managing each enterprise’s content portal.[65]

Though conventional approaches to content integration seem to lead to high per document set-up and maintenance costs, these should be contrasted with standard practice that suggests it may cost on average $25 to $40 per document simply for filing.²⁹ Indeed, labor costs can account for up to 30% of total document handling costs.²⁸ Nonetheless, at $5 to $11 per document for content management alone, this could result in no actual cost savings if electronic access does not displace current filing practices. When multiplied across all enterprise documents, these uncertainties can translate into huge swings in costs or benefits for a content portal initiative.

• Software License v. Full Project Costs – according to Charles Phillips of Morgan Stanley, only 30% of the money spent on major software projects goes to the actual purchase of commercially packaged software. Another third goes to internal software development by companies. The remaining 37% goes to third-party consultants.[66] In evaluating a commitment, internal staff and consulting time should be carefully scrutinized. Efficiencies in initial deployment and ongoing support are the biggest cost drivers
• Internal PLUS External Sources – weaknesses in scalability and high implementation costs often lead to a dismissal of the importance of integrating internal plus external content. Few installations address relevant content external to the enterprise essential to achieving its missions. Granted, the increase in scales associated with external content are large, but for some businesses integration with external content may be essential.

While other vendors claim fast categorization times, what they fail to mention is the lengthy pre-processing times necessary for generating their categorization metatags. According to Forrester Research, some of these metatagging systems can only process five to 15 documents per hour![67]

‘Cost’ of Inaccessible or Hidden Intranet Sites

In 2003, the portal vendor Plumtree noticed a new trend that it called “Web sprawl,” by which it meant the costly proliferation of Web applications, intranets and extranets.[68] BEA has taken up this trend as a major thrust to its Web service offerings through an approach it calls “enterprise portal rationalization” (EPR).[69] According to BEA, its architectural offerings are meant to control the “metastasizing” of corporate Web sites.

How common and to what scale is the proliferation of enterprise Web sites? I have not been able to find any comprehensive studies on this topic, but has been able to find many anecdotal examples. The proliferation, in fact, began as soon as the Internet became popular:

• As reported in 2000, Intel had more than 1 million URLs on its intranet with more than 100 new Web sites being introduced each month[70]
• In 2002, IBM consolidated over 8,000 intranet sites, 680 ‘major’ sites, 11 million Web pages and 5,600 domain names into what it calls the IBM Dynamic Workplaces, or W3 to employees[71]
• Silicon Graphics’ ‘Silicon Junction’ company-wide portal serves 7,200 employees with 144,000 Web pages consolidated from more than 800 internal Web sites[72]
• Hewlett-Packard Co., for example, has sliced the number of internal Web sites it runs from 4,700 (1,000 for employee training, 3,000 for HR) to 2,600, and it makes them all accessible from one home, @HP [73]^,[74]
• Avaya Corporation is now consolidating more than 800 internal Web sites globally[75]
• The Wall Street Journal recently reported that AT&T has 10 information architects on staff to maintain its 3,600 intranet sets that contain 1.5 million public Web pages[76]
• The new Department of Homeland Security is faced with the challenge of consolidating more than 3,000 databases inherited from its various constituent agencies.[77]

BrightPlanet’s customers confirm these trends, with indicators of hundreds if not thousands of internal Web sites common in the largest companies. Indeed, it is surprising how many instances there are where corporate IT does not even know the full extent of Web site proliferation. The problem is likely much greater than realized:

Table 13. Development and Unfound Document ‘Costs’ for Large Firms due to Web Sprawl

Table 13 consolidates previous information to estimate what the ‘costs’ of Web sprawl might be to larger firms (analogous to the Fortune 1000). The table presents Low, Medium and High estimates for number of Web sites per firm, known and unknown documents in each, and associated costs for initial site development and first-year maintenance plus the value of unfound information. The Medium category uses the average values from previous tables. The Low and High values bracket these amounts based on distribution of known values and expert judgment.

The table indicates as a mid-range estimate that an individual Web site for a large enterprise may cost about $6,000 to set-up and maintain in the first year and represents $24,000 in opportunity costs due to unknown or unfound documents. For the average large enterprise across all Web sites, these costs may be $4.2 million and $12.0 million, respectively. Across all large firms, total costs due to Web sprawl may be on the order of $22 billion.

While site development and maintenance costs are not trivial, exceeding $4 billion for all large firms (which can also be significantly reduced  – see previous section), the major cost impact comes from the inability to find or federate the information that is available. Unfound documents represent well in excess of 80% of the costs associated with Web sprawl.

The Web sprawl situation is analogous to other major technology shifts. For example, in the early 1980s, IT grappled mightily with the proliferation of personal computers. Centralized control was impossible in that circumstance because individuals and departments recognized the productivity benefits to be gained by PCs. Only when enterprise-capable vendors of networking technology, such as Novell, were able to offer integration solutions was the corporation able to control and fully exploit the PC’s technology potential.

The proliferation of internal enterprise Web sites is responding to similar drivers: innovation, customer service, or superior methods of product or solutions delivery. Ambitious mid-level managers will continue to exploit these advantages by “cowboy” additions of more corporate Web sites, and that is likely to the good for most enterprises. Gaining control and fully realizing the value of this Web site proliferation  – while not stymieing innovation  – will likely require enabling technology analogous to the networking of PCs.

IV. OPPORTUNITIES AND THREATS

The previous analysis has focused on more-or-less direct costs and drivers. These impacts are huge and deserve proper consideration. But there are other implications from the inability to access and manage relevant document information. These implications fall into the categories of lost opportunities, liabilities, or non-compliance. These implications often far outweigh the direct costs in their bottom-line impacts. This section presents only a few of these many opportunities.

‘Costs’ and Opportunity Costs of Winning Proposals

Competitive proposals are an important revenue factor to hundreds of thousands of businesses. Indeed, contracts and grants from federal, state and local governments accounted for 12.1% of GDP in 2002; the amount competitively awarded equaled about 5.6% of GDP.[78] Reducing the fully-burdened costs of producing responses to competitive procurements and improving the rate of successfully obtaining them can be a huge competitive advantage to business.

Significant proportions of commercial projects and programs are likewise awarded through competitive proposals and bids. However, literature references to these are limited, and the remainder of this section relies on federal sector statistics as a proxy for the overall category.

Though the federal government is making strides in providing central clearinghouses to opportunities  – and is also doing much in moving to uniform application standards and electronic application submissions  – these efforts are still in their nascent stages and similar efforts at the state and local level are severely lagging. As a result, the magnitude of the proposal opportunity is perhaps largely unknown to many businesses. This lack of appreciation and attention to the cost- and success-drivers behind winning proposals is a real gap in the competitiveness of many individual businesses.

Table 14 on the following page consolidates information from many government sources to quantify the magnitude of this competitively-awarded grant and contract opportunity with governments.

Table 14. Federal, State & Local Contract and Grant Opportunities, 2002

This analysis suggests there are nearly $600 billion available each year for competitively awarded grants and procurements from all levels of government within the U.S.; about 60% from the federal sector. The average competitive award is about $270 K for grants; about $220 K for contract procurements.

Aside from construction firms (which are excluded in this and prior analyses), there are on the order of 92,500 federal contract-seeking firms today.[87] In 2003, the top 200 federal contracting firms accounted for nearly $190 billion in contract outlays.[88] While it is unclear what proportion of these commitments were competitive (81% of total federal commitments) or based on all contract procurements (57% of total federal commitments), it is clear that more than 90,000 firms are competing via a classic power curve for a minor portion of available federal revenues. This power curve is shown in Figure 3 below for the 200 largest federal contractors, which obtain a proportionately high percentage of all contract dollars.

Figure 3. Power Curve Distribution of Top 200 Federal Contractors by Revenue, 2002

The combination of these factors enables an estimate of the bottom-line proposal impacts by firm. This information is shown in the table below:

Table 15. Combined Preparation Costs and Opportunity Costs for Proposals

Across all entities, the annual cost of preparing proposals to competitive solicitations from government agencies at all levels is on the order of $22 billion, $5 billion for winning firms and $17 billion for losing firms. Better access to missing information and better information  – assuming no change in the underlying ideas or proposal-writing skills  – suggests that proposal response costs could be reduced by more than $3 billion annually. Another $3 billion annually is available for better winning of competitive proposals. Individual benefits to firms that respond to competitive solicitations is on average $1.25 million per competing firm.[95]

The more significant benefit to individual firms from improved access to “missing” information and better information is increasing the likelihood of winning a competitive award. Firms that embrace these practices are estimated to obtain a $1.2 million annual benefit. Given that many firms that have previously been losing awards have relatively low annual revenues, the percent impact on the bottom line can be quite striking due to improved proposal preparation information.

‘Costs’ of Regulation and Regulatory Non-compliance

A December 2001 small business poll by the National Federation of Independent Business (NFIB) gauged the impacts of the regulatory workload on firms. When asked “is government regulation a very serious, somewhat serious, not too serious, or not at all serious problem for your business,” nearly half, or 43.6 percent, answered “very serious” or “somewhat serious.” The respondents indicated the most serious regulatory problems were at the federal level (49 %), state level (35 %) or local level (13%) of government. The biggest single regulatory problem cited was extra paperwork, followed by difficulty understanding how to comply with regulations and dollars spent doing so.[96] A later December 2003 NFIB survey indicates that the average cost per hour of complying with paperwork requirements was $48.72.[97]

Table 16. Per Employee Costs of Federal Regulation by Firm Size, 2002

According to a 2001 report, “The Impact of Regulatory Costs on Small Firms” by W. Mark Crain and Thomas D. Hopkins, the total costs of Federal regulations were estimated to be $843 billion in 2000, or 8 percent of the U. S. Gross Domestic Product. Of these costs, $497 billion fell on business and $346 billion fell on consumers or other governments. Here are how those impacts are estimated on a per employee basis across a range of firm sizes:[98]

As of September 30, 2002, federal agencies estimated there were about 8.2 billion “burden hours” of paperwork government-wide. Almost 95 percent of those 8.2 billion hours were being collected primarily for the purpose of regulatory compliance. [99]

Table 17. Federal Government Paperwork Burdens, 2002[100]

A December 2003 NFIB survey indicates that the average cost per hour of complying with paperwork requirements was $48.72.[101] If these costs are substituted, the total cost burden in the table above would be about $400 billion, $71 billion of which excludes Treasury and the IRS.

Despite legislation requiring federal paperwork reduction and embracing of e-government initiatives, paperwork burdens continue to increase. Total burden hours in 2002, for example, increased 600 million hours, or about 4 percent, from the previous year. The Code of Federal Regulations (CFR) continues to expand despite efforts to curtail further growth. The CFR grew from 71,000 pages in 1975 to 135,000 pages in 1998. Annually, there are more than 4,000 regulatory changes introduced by the federal government. The federal government now has over 8,000 separate information collection requests authorized by OMB.[102]

Table 18. Federal Fines and Penalties to Corporations, 2002

Another source of costs to enterprises are civil penalties and fines for non-compliance with existing regulations, as shown in the table above for 2002 by agency. A total of $5 billion annually is expended by U.S. businesses for civil penalties due to non-compliance with federal regulation, $1 billion of which is due to non-tax purposes.

However, these estimates may undercount actual fines and penalties levied by the federal government due to the accounting basis of the OMB source. For example, the Department of Labor (DOL) collected fines and penalties totaling $175 million from employers in fiscal year 2002 for Fair Labor Standards Act (FLSA) violations.[107] According to a 2002 report, since 1990, 43 of the government’s top contractors paid approximately $3.4 billion in fines/penalties, restitution, and settlements.[108] And, according to another report, the corporations liable to the top 100 False Claims Act paid more than $12 billion since 1986.[109] Since there is no central clearinghouse for this information, with both individual agency general counsels and the Department of Justice responsible for actual collections, the figures in Table 18 should be interpreted as estimates.

Table 19 on the next page consolidates the information in Table 16 to Table 18 to estimate the overall regulatory and paperwork burdens on U.S. businesses, plus estimates of the benefits to be gained from better document access and use.

‘Cost’ of an Unauthorized Posted Document

Unauthorized information disclosures derive mainly from within an organization. The ease of electronic record duplication and dissemination  – particularly through postings on enterprise Web sites  – increases a firm’s vulnerability to this problem. Records mutate and propagate in poorly controlled environments. On average, unauthorized disclosure of confidential information costs Fortune 1000 companies about $15 million per company per year.[110]

A few privacy laws demonstrate the potential liabilities associated with disclosure of confidential information due to inadvertent mistakes or disgruntled employees. As one example, the Health Insurance Portability and Accountability Act (HIPAA) of 1996 sets security standards protecting the confidentiality and integrity of “individually identifiable health information,” past, present or future. Failure to comply with any of the electronic data, security, or privacy standards can result in civil monetary penalties up to $25,000 per standard per year. Violation of the privacy regulations for commercial or malicious purposes can result in criminal penalties of $50,000 to $250,000 in fines and one to ten years of imprisonment.[111]

Table 19. Regulatory Burden and Benefits to Firms from Improved Information

As another example, the Gramm-Leach-Bliley Act (GLBA) of 1999 mandates the financial industry to create guidelines for the safeguarding of customer information. GLBA includes severe civil and criminal penalties for non-compliance, with civil penalties up to $100,000 for each violation and key officers may be fined up to $10,000 per violation. Violation of the GLBA can also carry hefty sanctions, including termination of FDIC insurance and fines of up to $1,000,000 for an individual or one percent of the total assets of the financial institution.[117]

Other major areas of unauthorized disclosure liability occur in national security, identity theft, and commerce, tax and Social Security information. Indeed, virtually every state and federal agency related to a company’s business has policies and fines regarding unauthorized disclosures. Monitoring these requirements is thus an imperative for enterprise management to prevent exposure to fines and loss of reputation.

On a less-quantifiable basis there are also risks about the clarity of the enterprise message to customers, suppliers and partners. Unmanaged Web sprawl is a critical hole for enterprises to ensure compliance with privacy and confidentiality regulations, and to promote clarity of message and accuracy to stakeholders.

V. CONCLUSIONS

Prior to the analysis in this white paper, the state of understanding about the value of document assets had been abysmal. While still preliminary and subject to much improvement, this study has nonetheless found:

• The value of documents  – in their creation, access and use  – can indeed be measured
• The information contained within U.S. enterprise documents represents about a third of gross domestic product, or an amount of about $3.3 trillion annually
• Some 25% of all of these expenditures lend themselves to actionable improvements
• There are perhaps on the order of 10 billion documents created annually in the U.S.
• Corporate data doubles every six to eight months; 85% of this data is contained in documents
• Ninety to 97 percent of enterprises cannot estimate how much they spend on producing documents each year
• Document creation is about 2-3 times more important  – from an embedded cost standpoint  – than document handling
• It costs, on average, $350 to create a ‘typical’ document
• The total potential benefit from practical improvements in document access and use to the U.S economy is on the order of $800 billion annually, or about 8% of GDP
• For the 1,000 largest U.S. firms, benefits from these improvements can approach nearly $250 million annually per firm
• About three-quarters of these benefits arise from not re-creating the intellectual capital already invested in prior document creation
• Another 25% of the benefits are due to reduced regulatory non-compliance or paperwork, or better competitiveness in obtaining solicited contracts and grants
• $33 billion is wasted each year in re-finding previously found Web documents
• Paperwork and regulatory improvements due to documents can save U.S. enterprises $120 billion each year
• Lack of document access due to Web sprawl costs U.S. enterprises $22 billion each year
• $8 billion in annual benefits is available due to document improvements for competitive governmental grant and contract solicitations
• These figures likely severely underestimate the benefits to enterprises from improved competitiveness, a factor not analyzed in this study
• Documents are now at the point where structured data was at 15 years ago at the nascent emergence of the data warehousing market.

As noted throughout, there is a considerable need for additional research and data on document creation, use, costs and benefits. Additional technical endnotes are provided in the PDF version of the full paper.

[1] All sources and assumptions are fully documented in footnotes in the main body of this white paper; general assumptions used in multiple tables are provided in the Technical Endnotes.

[2] As quoted by Armando Garcia, vice president of content management at IBM; see http://www.contentworld.com/conference/conthur.html

[3] Delphi Group, “Taxonomy & Content Classification Market Milestone Report,” Delphi Group White Paper, 2002. See http://delphigroup.com.

[4] Based on the 1999 to 2001 estimate changes in reference 34, Table 2-6.

[5] As initially published in Inc Magazine in 1993. Reference to this document may be found at: http://www.contingencyplanning.com/PastIssues/marapr2001/6.asp

[6] J. Snowdon, Documents – The Lifeblood of Your Business?, October 2003, 12 pp. The white paper may be found at: http://www.mdy.com/News&Events/Newsletter/IDCDocMgmt.pdf

[7] Xerox Global Services, Documents – An Opportunity for Cost Control and Business Transformation, 28 pp., 2003. The findings may be found at: http://www.sap.com/solutions/srm/pdf/CCS_Xerox.pdf

[8] A.T. Kearney, Network Publishing: Creating Value Through Digital Content, A.T. Kearney White Paper, April 2001, 32 pp. See http://www.adobe.com/aboutadobe/pressroom/pressmaterials/networkpublishing/pdfs/netpubwh.pdf.

[9] S.A. Mohrman and D.L. Finegold, Strategies for the Knowledge Economy: From Rhetoric to Reality, 2000,http://www.marshall.usc.edu/ceo/Books/pdf/knowledge_economy.pdf. University of Southern California study as supported by Korn/Ferry International, January 2000, 43 pp. See

[10] C. Moore, TheContent Integration Imperative, Forrester Research Trends Report, March 26, 2004, 14 pp.

[11] D. Vesset, Worldwide Business Intelligence Forecast and Anal ysis, 2003-2007, International Data Corporation, June 2003, 18 pp. See http://www.dwway.com/file/20030708085453_IDC_WW-BIFORECASTANDANALYSIS2003-07_JUN03.pdf.

[12] M. Stonebraker and J. Hellerstein, “Content Integration for E-Business,” in ACM SIGMOD Proceedings, Santa Barbara, CA, pp. 552-560, May 2001.

[13] P. Lyman and H. Varian, “How Much Information, 2003,” retrieved from http://www.sims.berkeley.edu/how-much-info-2003 on December 1, 2003.

[14] U.S. Department of Commerce, Digital Economy 2003, Economic Statistics Administration, U.S. Dept. of Commerce, Washington, D.C., April 2004, 155 pp. See http://www.esa.doc.gov/DigitalEconomy2003.cfm.

[15] U.S. Department of Labor, “Occupation Employment and Wages, 2002,” Bureau of Labor Statistics. See http://www.bls.gov/news.release/archives/ocwage_11192003.pdf.

[16] U.S. Census Bureau, “Statistics of U.S. Businesses 2001.” See http://www.census.gov/epcd/susb/2001/us/US–.htm.

[17] Total office documents counts were obtained on a page basis from reference 13, which used a value of 2% for what documents deserve to be archived. This formed the ‘lo’ case, with the high case using a 5% estimate (lower still than the ENST 10% estimated cited in reference 13). Total pages were converted to numbers of documents on an average 8 pp per document basis; see Technical Endnotes for further discussion.

[18] See Technical Endnotes for the derivation of knowledge worker estimates.

[19] See Technical Endnotes for the derivation of content worker estimates.

[20] Citation sources and assumptions for this analysis are presented in the BrightPlanet white paper, “A Cure to IT Indigestion: Deep Content Federation,” BrightPlanet Corporation White Paper, June 2004, 31 pp.

[21] The “bottom up” cases are built from the number of assumed knowledge workers in Table 3. The “low” and “high” variants are based on a 5% archival value or 350 annual documents created per worker, respectively, applied to worker staff costs associated with document creation. The “Coopers & Lybrand” case is a strict updating of that study to 2002. The other two “C&L” cases use the updated per document costs from the C&L study; the first variant uses the annual documents created from the UC Berkeley study without archiving; the second variant uses the average of the “low” and “high” document numbers. See further Technical Endnotes for other key assumptions.

[22] The individual values in Table 5 range from about $140 to $740 per document, with the update of the Coopers & Lybrand study being about $270. Separate Delphi analysis by BrightPlanet has shown median values of about $550 per document.

[23] See http:// www.eds.com/services_offerings/ibill_openbill_b2b.shtml

[24] See http://www.hsh.com/cfee-sample.html.

[25] See http://www.atp.nist.gov/eao/applicants/section9.htm.

[26] As initially published in Inc Magazine in 1993. Reference to this document may be found at: http://www.contingencyplanning.com/PastIssues/marapr2001/6.asp

[27] Xerox Global Services, Documents – An Opportunity for Cost Control and Business Transformation, 28 pp., 2003. The findings may be found at: http://www.sap.com/solutions/srm/pdf/CCS_Xerox.pdf and J. Snowdon, Documents  – The Lifeblood of Your Business?, October 2003, 12 pp. The white paper may be found at: http://www.mdy.com/News&Events/Newsletter/IDCDocMgmt.pdf

[28] Optika Corporation. See http://www.optika.com/ROI/calculator/ROI_roiresults.cfm.

[29] Cap Ventures information, as cited in ZyLAB Technologies B.V., “Know the Cost of Filing Your Paper Documents,” Zylab White Paper, 2001. See http://www.zylab.com/downloads/whitepapers/PDF/21%20-%20Know%20the%20cost%20of%20filing%20your%20paper%20documents.pdf.

[30] ALL Associates Group, Inc., EDAM Sector Summary, April 2003, 2 pp.

[31] ALL Associates Group, 2002 EDAM Metrics for Major U.S. Companies.

[32] By the second Q 2004, this amount was $11.6 trillion. U.S. Federal Reserve Board, Flow of Funds Accounts for the United States, Sept. 16, 2004. See http://www.federalreserve.gov/releases/Z1/current/accessible/f6.htm.

[33] The bases for this table have the following assumptions: 1) the three cases for document handling are based on 5%, 10% and 15% of total enterprise revenues, per the earlier section; 2) the three cases for document creation are based on the ‘C&L Bottom-Up’, ‘Bottom-up  – High,’ and ‘Coopers & Lybrand’ items for the Low, Medium, and High columns, respectively, in Table 5; and 3) the document misfiling case draws on the same basis but using the total document estimates and misfiled percentages of 5%, 7.5% and 9% consistent with the previous discussion section. See further the Technical Endnotes.

[34] P. Lyman and H. Varian, “How Much Information, 2003,” retrieved from http://www.sims.berkeley.edu/how-much-info-2003 on December 1, 2003.

[35] Cap Ventures information, as cited in ZyLAB Technologies B.V., “Know the Cost of Filing Your Paper Documents,” Zylab White Paper, 2001. See http://www.zylab.com/downloads/whitepapers/PDF/21%20-%20Know%20the%20cost%20of%20filing%20your%20paper%20documents.pdf.

[36] As reported in http://www.hoovers.com/company/archive/detail/0,2049,7_2322,00.html.

[37] See http://www.veronissuhler.com/businfo/segment.html, August 2, 2000.

[38] See http://www.outsellinc.com/docs/pr_release/pr20000602_01.htm, June 2, 2000.

[39] See http://www.outsellinc.com/docs/pr_release/pr20000629_01.htm.

[40] M.K. Bergman, “The Deep Web: Surfacing Hidden Value,” BrightPlanet Corporation White Paper, June 2000. The most recent version of the study was published by the University of Michigan’s Journal of Electronic Publishing in July 2001. See http://www.press.umich.edu/jep/07-01/bergman.html.

[41] This analysis assumes there were 1 million documents on the Web as of mid-1994.

[42] See, for example, C. Sherman and G. Price, The Invisible Web, Information Today, Inc., Medford, NJ, 2001, 439 pp., and P. Pedley, The Invisible Web: Searching the Hidden Parts of the Internet, Aslib-IMI, London, 2001, 138pp.

[43] iProspect Corporation, iProspect Search Engine User Attitudes, April/May 2004, 28 pp. See http://www.iprospect.com/premiumPDFs/iProspectSurveyComplete.pdf.

[44] As reported at http://www.nua.ie/surveys/index.cgi?f=VS&art_id=905358569&rel=true.

[45] Delphi Group, “Taxonomy & Content Classification Market Milestone Report,” Delphi Group White Paper, 2002. See http://delphigroup.com.

[46] C. Sherman and S. Feldman, “The High Cost of Not Finding Information,” International Data Corporation Report #29127, 11 pp., April 2003.

[47] M.E.D. Koenig, “Time Saved  – a Misleading Justification for KM,” KMWorld Magazine, Vol 11, Issue 5, May 2002. See http://www.kmworld.com/publications/magazine/index.cfm.

[48] G. Xu, A. Cockburn and B. McKenzie, Lost on the Web: An Introduction to Web Navigation Research, http://www.cosc.canterbury.ac.nzq/ACMchapterq/NZCSPGq/papers.

[49] A. Cockburn and B. McKenzie, What Do Web Users Do? An Empirical Analysis of Web Use, 2000. See http://citeseer.ist.psu.edu/cockburn00what.html.

[50] Tenth edition of GVU’s (graphics, visualization and usability} WWW User Survey, May 14, 1999. See http://www.gvu.gatech.edu/user_surveys/survey-1998-10/tenthreport.html.

[51] C. Alvarado, J. Teevan, M. S. Ackerman and D.Karger, “Surviving the Information Explosion: How People Find Their Electronic Information,” AI Memo 2003-06, April 2003, 11 pp.., Massachusetts Institute of Technology, Computer Science and Artificial Intelligence Laboratory. See ftp://publications.ai.mit.edu/ai-publications/2003/AIM-2003-006.pdf.

[52] W. Jones, H. Bruce and S. Dumais, “Keeping Found Things Found on the Web,” See http://washington.edu/KFTF_Web.pdf.

[53] J. Teevan, “How People Re-find Information When the Web Changes,” AI Memo 2004-014, June 2004, 10 pp., Massachusetts Institute of Technology, Computer Science and Artificial Intelligence Laboratory. See ftp://publications.ai.mit.edu/ai-publications/2004/AIM-2004-012.pdf.

[54] Library of Congress, “Preserving Our Digital Heritage: Plan for the National Digital Information Infrastructure and Preservation Program”, a Report to Congress by the U.S. Library of Congress, 2002, 66 pp. See http://www.digitalpreservation.gov/ndiipp/.

[55] Consistent with Table 8; this analysis also assumes the 25% search time commitment by employee and previous values from earlier tables.

[56] All subsequent references to ‘Large’ firms is based on the last column in Table 2, namely the 930 U.S. firms with more than 10,000 employees.

[57] Delphi Group, “Taxonomy & Content Classification Market Milestone Report,” Delphi Group White Paper, 2002. See http://delphigroup.com.

[58] S. Stearns, “Realize the Value Locked in Your Content Silos Without Breaking the Bank: Automated Classification Tools to Improve Information Discovery,” Inmagic White Paper, version 1.0, 2004. 10 pp. See http://www.inmagic.com.

[59] P. Sonderegger, “Weave Search into the Browsing Experience,” ForresterQuick Take, Forrester Research, Inc., Feb. 18, 2004. 2 pp.

[60] P. Russom, “An Eye for the Needle,” Intelligent Enterprise, January 14, 2002. See http://www.iemagazine.com/020114/502feat2_1.

[61] This average was estimated by interpolating figures shown on Figure 8 in reference 68.

[62] This average was estimated by interpolating figures shown on the p.14 figure in Plumtree Corporation, “The Corporate Portal Market in 2002,” Plumtree Corp. White Paper, 27 pp. See http://www.plumtree.com/pdf/Corporate_Portal_Survey_White_Paper_February2002.pdf.

[63] The ‘low’ case represents the archival value in the middle bars with the addition that 30% of internal documents generated in the current year have a value to be shared for one year; the ‘high’ case represents the related archival value in the middle bars but with 40% of documents generated in that year having a value to be shared for one year.

[64] Analysis based on reference 68, with interpolations from Figure 16.

[65] M. Corcoran, “When Worlds Collide: Who Really Owns the Content,” AIIM Conference, New York, NY, March 10, 2004. See http://show.aiimexpo.com/convdata/aiim2003/brochures/64CorcoranMary.pdf.

[66] C. Phillips, “Stemming the Software Spending Spree,” Optimize Magazine, April 2002, Issue 6. See http://www.optimizemag.com/article/showArticle.jhtml?articleId=17700698&pgno=1.

[67] C. Moore, “The Content Integration Imperative,” Forrester Research, Inc., March 26, 2004, 14 pp.

[68] Plumtree Corporation, “The Corporate Portal Market in 2003,” Plumtree Corp. White Paper, 30 pp. See http://www.plumtree.com/portalmarket2003/default.asp.

[69] BEA Corporation, “Enterprise Portal Rationalization,” BEA Technical White Paper, 23 pp., 2004. See http://www.bea.com/content/news_events/white_papers/BEA_epr_wp.pdf.

[70] A. Aneja, C.Rowan and B. Brooksby, “Corporate Portal Framework for Transforming Content Chaos on Intranets,” Intel Technology Journal Q1, 2000. See http://developer.intel.com/technology/itj/q12000/pdf/portal.pdf.

[71] J. Smeaton, “IBM’s Own Intranet: Saving Big Blue Millions,” Intranet Journal, Sept. 25, 2002. See http://www.intranetjournal.com/articles/200209/ij_09_25_02a.html.

[72] See http://www.wookieweb.com/Intranet/.

[73] D. Voth, “Why Enterprise Portals are the Next Big Thing,” LTI Magazine, October 1, 2002. See http://www.ltimagazine.com/ltimagazine/article/articleDetail.jsp?id=36877.

[74] A. Nyberg, “Is Everybody Happy?” CFO Magazine, November 01, 2002. See http://www.cfo.com/article/1%2C5309%2C8062%2C00.html.

[75] See http://www.proudfoot-plc.com/pdf_20004-USPR1002Avayaweb.asp.

[76] Wall Street Journal, May 4, 2004, p. B1.

[77] pers. comm.., Jonathon Houk, Director of DHS IIAP Program, November 2003.

[78] These figures are based on Table 12 and the GDP figures from reference 32. Note, the analysis in this section also ignores business-to-business opportunities, which are also likely significant.

[79] Total grant and procurement amounts are derived from the U.S. Census Bureau, Consolidated Federal Funds Report (CFFR). See http://harvester.census.gov/cffr/asp/Reports.asp.

[80] The number of awards and an analysis of which line items are competitively awarded was derived from the U.S. Census Bureau, Federal Assistance Award Data System (FAADS). See http://www.census.gov/govs/faads/021sumus.htm.

[81] Specific categories of grants were analyzed based on the U.S. General Services Administration’s Catalog of Federal Domestic Assistance (CFDA) definitions to determine degree of competitiveness; see http://12.46.245.173/cfda/cfda.html. Figures from the U.S. Department of Health and Human Services, Grant.gov Clearinghouse (see http://www.grants.gov/) suggest that $350 billion in federal grants is available, but many of the specific grant opportunities are geared to state governments or individuals. That is why the figures shown indicate only $100 billion in competitive opportunities available directly to enterprises.

[82] U.S. General Services Administration, Federal Procurement Data System  – NG (FY 2003 data); see http://www.fpdc.gov/fpdc/FPR2003a.pdf and http://www.fpdc.gov/fpdc/FPR2003c.pdf. These sources are also the reference for the number of actions or successful awards. Due to discrepancies, these amounts were adjusted to conform with the totals in reference 79.

[83] Average competitive opportunities are derived by dividing the total award amount by category by the number of awards for that category.

[84] See http://www.gcswin.com/opportunities/opp2.htm. This is the only summary reference for state and local information found. Splits between grants and contract procurements were adjusted based on the assumption that contract amounts differed at the non-federal level. Thus, while the split for grant-contract procurements in the federal sector is about 58%-42% in the federal sector, it is assumed to be 38%-62% at the state and local level.

[85] There may also be some double counting of state amounts due to transfers from the federal government. For example, in 2002, $360,534 million in direct transfers was made to states and localities from the federal government. U.S. Census Bureau, State and Local Government Finances by Level of Government and by State: 2001  – 02. See http://www.census.gov/govs/estimate/0200ussl_1.html.

[86] This analysis assumes that individual grant and contract awards are 80% of the amount shown at the federal level.

[87] To be listed requires a minimum of $10,000 in federal contracts; see http://clinton2.nara.gov/WH/EOP/OP/html/aa/aa06.html.

[88] See http://www.govexec.com/features/0804-15/0804-15s1s1.htm.

[89] This header information is drawn from Table 12.

[90] Number of competing firms is increased from the federal contractor baseline by a factor of 1.30 to account for new state and local government contractors.

[91] Winning and losing proposal preparation costs are based on the empirical percentages from NIST (see reference 93), namely 0.85% and 0.59%, respectively, as a percent of total award amounts.

[92] The ‘Low’ basis for improvements is based on the finding of missing information discussed in a previous section; the ‘High” basis reflects the difference between lowest quartile and highest quartile efforts spent on successful proposal preparation (see reference 93). The ‘Med’ basis is an intermediate value between these two.

[93] The increase in winning submissions is calculated based on numbers of winning proposals times the RFP improvement factor. In fact, because all things being equal the pool of contract dollars does not change, this amount merely represents a shift of winning awards from existing winners to new winners. In other words, total contracts amounts are a zero-sum game with proposal improvements by previous losers taken from the pool of previous winners.

[94] The analysis in Figure 2 indicates there is a power curve distribution of awards. The number of new winning proposals was applied to this curve to estimate the actual number of new firms winning awards; see Figure 2 for the power-curve fitting equation.

[95] Of course, better probabilities of winning competitive solicitations are a zero-sum game. New winners displace old winners. The real advantage in this arena is to individual firms that better succeed at securing the existing pool of competitive funds. The benefits to individual companies can be the difference between profitability, indeed survival.

[96] NFIB, Coping with Regulation, NFIB National Small Business Poll, Vol. 1, Issue 5. See http://www.nfib.com/object/3105105.html.

[97] NFIB, Paperwork and Record-keeping, NFIB National Small Business Poll, Vol. 3, Issue 5. See http://www.nfib.com/object/4131277.html.

[98] W. M. Crain & T. D. Hopkins, “The Impact of Regulatory Costs on Small Firms”, Report to the Small Business Administration, RFP No. SBAHQ-00-R-0027 (2001). The report’s 2000 year basis was updated to 2002 based on a 4% annual inflation factor.

[99] U.S. General Accounting Office, Paperwork Reduction Act: Record Increase in Agencies’ Burden Estimates, testimony of V. S. Rezendes, before the Subcommittee on Energy, Policy, Natural Resources and Regulatory Affairs, Committee on Government Reform, House of Representatives, April 11, 2003. See http://www.reform.house.gov/UploadedFiles/Testimony_GAO_Revised.pdf.

[100] Office of Management and Budget, Managing Information Collection and Dissemination, Fiscal Year 2003, 198 pp. (Table A1). See http://www.whitehouse.gov/omb/inforeg/2003_info_coll_dism.pdf.

[101] NFIB, Paperwork and Record-keeping, NFIB National Small Business Poll, Vol. 3, Issue 5. See http://www.nfib.com/object/4131277.html.

[102]U.S. Small Business Administration, Final Report of the Small Business Paperwork Relief Task Force, June 27, 2003, 64 pp. See http://www.sbaonline.sba.gov/advo/laws/final_paperwork03.pdf.

[103] IRS, Civil Penalties Assessed and Abated, by Type of Penalty and Type of Tax (Table 26), September 20, 2002. See http://www.irs.gov/pub/irs-soi/02db26cp.xls.

[104] Except as footnoted, the figures below are drawn from the OMB Public Budget Tables. Civil penalties for crime victims have been excluded from these figures. See http://www.whitehouse.gov/omb/budget/fy2005/db.html.

[105] Obtained orders in SEC judicial and administrative proceedings requiring securities law violators to disgorge illegal profits of approximately $1.293 billion. Civil penalties ordered in SEC proceedings totaled approximately $101 million. See SEC http://www.sec.gov/pdf/annrep02/ar02enforce.pdf.

[106] T. L. Sansonetti, U.S. Department of Justice, testimony before the House Committee on the Judiciary, Subcommittee on Commercial and Administrative Law, March 9, 2004. See http://www.house.gov/judiciary/sansonetti030904.htm.

[107]Argy, Wiltse & Robinson, Business Insights, Summer 2003, 4 pp. See http://www.awr.com/news_let/Argy%20Summer%202003.pdf

[108] Project on Government Oversight, Federal Contractor Misconduct: Failures of the Suspension and Debarment System, revised May 10, 2002. See http://www.pogo.org/p/contracts/co-020505-contractors.html.

[109]Corporate Crime Reporter, Top 100 False Claims Act Settlements, December 30, 2003, 64 pp. See http://www.corporatecrimereporter.com/fraudrep.pdf.

[110] According to Alchemia Corporation testimony citing a Price Waterhouse Coopers study, FDA Hearing, Jan. 17, 2002. See http://www.fda.gov/ohrms/dockets/dockets/ 00d1538/00d-1538_mm00023_01_vol7.doc.

[111] For example, see http://www.medschool.ucsf.edu/curriculum/clinical/guide/section2/confidentiality.asp.

[112] From Table 17.

[113] From Table 16 after adjusting by total number of employees for all firms as shown on Table 2, and removal of total burdens as shown in Table 17.

[114] From Table 18.

[115] All ‘State and Local’ items are based on the ratio of state and local budgets in relation to the federal budget, excluding direct federal transfers, and applied to those factors for the federal sector. This ratio is 0.563. See http://www.gpoaccess.gov/usbudget/fy01/guide01.html.

[116] All ‘Large Firm’ estimates are based on the ratio of large firm documents to total firm documents; see Table 2.

[117] For example, see http://www.nfr.com/why/mandates.php#gramm

I just came across a VC blog pondering the value to a start-up of operating in “Stealth Mode” or not.  I’ve amusingly come to the conclusion that all of this — particularly the “stealth” giveaway — is so much marketing hype.  When a start-up claims they’re coming out of stealth mode, grab your wallet.

The most interesting and telling example I have of this is Rearden Commerce, which was announced in a breathy cover story in InfoWorld in February 2005 about the company and its founder/CEO Patrick Grady.  The company has an obvious “in” with the magazine; in 2001 InfoWorld also carried a similar piece on the predecessor company to Rearden, Talaris Corporaton.

According to a recent Business Week article, Rearden Commerce and its predecessors reaching back to an earlier company called Gazoo founded in 1999 have raised $67 million in venture capital.  While it is laudable the founder has reportedly put his own money into the venture, this venture through its massive funding and high-water mark of 80 employees or so hardly qualifies as “stealth.”

As early as 2001 with the same technology and business model, this same firm was pushing the “stealth” moniker. According to an October 2001 press release:

“The company, under its stealth name Gazoo, was selected by Red Herring magazine as one of its ‘Ten to Watch’ in 2001.”  [emphasis added]

Even today, though no longer using the active name Talaris Corporation, it has close to 115,000 citations on Yahoo! Notable VCs such as Charter Ventures, Foundation Capital, JAFCo and Empire Capital have backed it through its multiple incubations.

Holmes Report, a marketing company, provides some insight into how the earlier Talaris was spun in 2001:

“The goal of the Talaris launch was to gain mindshare among key business and IT trade press and position Talaris as a ‘different kind of start-up’ with a multi-tiered business model, seasoned executive team and tested product offering.”

[Hmmm; grind me a pound!]

The Holmes Report documents the analyst firms and leading journals and newspapers to which it made outreach.  Actually, this outreach is pretty impressive.  Good companies do the same all of the time and that is to be lauded.  What is to be questioned, however, is how many “stealths” a cat can have.  Methinks this one is one too many.

“Stealth” thus appears to be code for an existing company of some duration that has had disappointing traction and now has new financing, a new name, new positioning, or all of the above.  So, interested in a start-up that just came out of stealth mode?  Let me humbly suggest standard due diligence.

This Friday brown bag leftover was first placed into the AI3 refrigerator on October 13, 2005. No changes have been made to the original posting, except the [grinding] bit.

However, as of last year, Rearden had upped its VC funding to $240 million (can we spell multiple ?). Today, it is now focused on the travel industry. Fly me to the moon!

(And Wishing All a Healthy and Prosperous 2010!)

According to iProspect, about 56 percent of users use search engines every day, based on a population of which more than 70 percent use the Internet more than 10 hours per week.[1] The average knowledge worker spends 2.3 hrs per day — or about 25% of work time — searching for critical job information.[2] IDC estimates that enterprises employing 1,000 knowledge workers may waste well over $6 million per year each in searching for information that does not exist, failing to find information that does, or recreating information that could have been found but was not.[3]

Vendors and customers often use time savings by knowledge workers as a key rationale for justifying a document or content initiative. This comes about because many studies over the years have noted that white collar employees spend a consistent 20% to 25% of their time seeking information. The premise is that more effective search will save time and drop these percentages. For example, EDS has suggested that improvements of 50 percent in the time spent searching for data can be achieved through improved consolidation and access to data.[4]

Using these premises, consultants often calculate that every 1% reduction in the total work time devoted to search works out illustratively on a fully burdened basis as a big cost savings benefit:

$50,000 (base salary) * 1.8 (burden rate) * 1.0% = $900/ employee

Beware such facile analysis!

The fact that many studies over the years have noted white collar employees spend a consistent 20% to 25% of their time devoted to search suggests it is the “satisficing” allocation of time to information search. (In other words, knowledge workers are willing to devote a quarter of their time to finding relevant information; the remainder for analysis and documentation.)

Thus, while better tools to aid better discovery may lead to finding better information and making better decisions more productively — an important justification in itself — there may not result a strict time or labor savings from more efficient search.[5] Be careful of justifying project expenditures based on “time savings” related to search. Search is likely to remain the “25% solution.” The more relevant question is whether the time that is spent on search produces better information or not.

This Friday brown bag leftover was first placed into the AI3 refrigerator on September 14, 2005. No changes have been made to the original posting.

Pinheads Sometimes Get the Last Laugh

The idea of the ‘long tail’ was brilliant, and Chris Anderson’s meme has become part of our current lingo in record time. The long tail is the colloquial name for a common feature in some statistical distributions where an initial high-amplitude peak within a population distribution is followed by a rapid decline and then a relatively stable, declining low-amplitude population that “tails off.” (An asymptotic curve.) This sounds fancy; it really is not. It simply means that a very few things are very popular or topical, most everything else is not.

The following graph is a typical depiction of such a statistical distribution with the long tail shown in yellow. Such distributions often go by the names of power laws, Zipf distributions, Pareto distributions or general Lévy distributions. (Generally, such curves when plotted on a semi-logarithmic scale now show the curve to be straight, with the slope being an expression of its “power”.)

It is a common observation that virtually everything measurable on the Internet — site popularity, site traffic, ad revenues, tag frequencies on del.icio.us, open source downloads by title, Web sites chosen to be digg‘ed, Google search terms — follows such power laws or curves.

However, the real argument that Anderson made first in Wired magazine and then in his 2006 book, The Long Tail: Why the Future of Business is Selling Less of More, is that the Internet with either electronic or distributed fulfillment means that the cumulative provision of items in the long tail is now enabling the economics of some companies to move from “mass” commodities to “specialized” desires. Or, more simply put: There is money to be made in catering to individualized tastes.

I, too, agree with this argument, and it is a brilliant recognition of the fact that the Internet changes everything.

But Long Tails Have ‘Teeny Heads’

Yet what is amazing about this observation of long tails on the Internet has been the total lack of discussion of its natural reciprocal: namely, long tails have teeny heads, the red portion of the diagram. For, after all, what also is the curve above telling us? While Anderson’s point that Amazon can carry millions of book titles and still make a profit by only selling a few of each, what is going on at the other end of the curve — the head end of the curve?

Well, if we’re thinking about book sales, we can make the natural and expected observation that the head end of the curve represents sales of the best seller books; that is, all of those things in the old 80-20 world that is now being blown away with the long tail economics of the Internet. Given today’s understandings, this observation is pretty prosaic since it forms the basis of Anderson’s new long tail argument. Pre-Internet limits (it’s almost like saying before the Industrial Revolution) kept diversity low and choices few.

Okaaaay! Now that seems to make sense. But aren’t we still missing something? Indeed we are.

Social Collaboration Depends on ‘Teeny Heads’

So, when we look at many of those aspects that make up what is known as Web 2.0 or even the emerging semantic Web, we see that collaboration and user-submitted content stands at the fore. And our general power law curves then also affirm that it is a very few who supply most of that user-generated content — namely, those at the head end, the teeny heads. If those relative few individuals are not motivated, the engine that drives the social content stalls and stutters. Successful social collaboration sites are the ones that are able to marshal “large numbers of the small percentage.”

The natural question thus arises: What makes those “teeny heads” want to contribute? And what makes it so they want to contribute big — that is, frequently and with dedication? So, suddenly now, here’s a new success factor: to be successful as a collaboration site, you must appeal to the top 1% of users. They will drive your content generation. They are the ‘teeny heads’ at the top of your power curve.

Well, things just got really more difficult. We need tools, mindshare and other intangibles to attract the “1%” that will actually generate our site’s content. But we also need easy frameworks and interfaces for the general Internet population to live comfortably within the long tail.

So, heads or tails? Naahh, that’s the wrong question. Keep flipping until you get both!

This Friday brown bag leftover was first placed into the AI3 refrigerator on April 6, 2007. I remain convinced that only a few “movers” are required to start shaking the user-generated content tree. I made two changes from the original post: 1) I had to find a grainy replacement for the initial great graphic of the “teeny head”; and 2) I added the red portion to the graphic. No other changes have been made.