New books for knowledge base publishers

July, 2003

This article discusses five recent books of interest to Knowledge Base Publishers on such topics as how to:

• add value to corporate intellectual assets;
• make Web sites easier to use;
• reduce costs and time delays associated with Internet publishing;
• organize and integrate information from multiple applications;
• improve performance and add value to search engines.

What makes these titles interesting is that they cover many of the same issues from five different perspectives:

1. corporate and project management (Unlocking Knowledge Assets by Conway and Sligar)
2. libraries and information services (Information Architecture by Rosenfeld and Morville)
3. corporate communications (Enterprise Content Services by Laugero and Globe)
4. enterprise systems development (Metadata Solutions by Tannenbaum)
5. IT specialist or "power user" (Google Hacks by Calishain and Dornfest)

For each book, we provide a description, highlights, and links to sample chapters.

1. Unlocking Knowledge Assets
(Susan Conway and Char Sligar, Microsoft Press, 2002)
This book, written by two Knowledge Management program managers at Microsoft, takes a high level, strategic approach supported by details and examples. The combination makes the book especially useful for developing a project plan.

The authors define KM as "moving information from one person to another in a reusable format" — a broad concept that includes personal, community, public, and corporate information spaces. Much of the material will be familiar to people who have reviewed the KM literature (see CFO's Guide to Intellectual Capital). Fresh material includes:

• a chapter on measuring the value of knowledge, including a "KM Value Assessment" model developed by Baruch Lev;
• a thorough discussion of technology, from KM architecture to repositories and search tools;
• a chapter on building taxonomies.

The book is refreshingly free of consulting jargon and marketing hype about Microsoft products. You get the big picture, but it's well grounded with examples, technical details, and implementation advice. A glossary of KM terms is included.

Highlights
Some of my favorite tidbits are the following:

• a chart, "Sample behaviors mapped to sample activities and performance goals," that illustrates how to develop performance yardsticks for KM projects in different corporate departments;
• an "Implementation Framework" that assigns a key role to editors who select content for specific audiences;
• the "Knowledge Index" — a hierarchical map showing how specific "knowledge assets" relate to different business processes and sub-processes, such as vendor selection and procurement in a supply chain function;
• the "community content" pyramid, which shows raw, invalidated knowledge assets at the base and fully validated best practices at the apex;
• the "Library" (KM repository) showing the relationship of content to users, metadata, the storage system, and program interfaces.

A sample chapter of this book,"Building taxonomies," is available on the Web.

2. Information architecture
(Louis Rosenfeld and Peter Morville, 2nd edition, O'Reilly, 2002)
The first edition of this book, written by two University of Michigan librarians, appeared in 1998 just as corporations began looking for alternatives to help users find information on large, complex Web sites. The authors' answer, which they call "information architecture," grew out of the orderly and familiar library paradigm with its directories, labeling systems, and subject headings.

The first edition succeeded partly because the concept filled a gap for corporate webmasters and partly because the authors have a knack for explaining arcane library science topics like controlled vocabularies, authority files, labeling systems, and metadata to the layman (see Mark Bernstein's review). In the second edition, the authors update the sections on organization, labeling, navigation, and searching. In the methodology chapters, they venture beyond library science to include a more interdisciplinary collection of tools and techniques (see our 1997 article "What is knowledge base publishing") and add more bottom-up strategies (see our 1999 article, "Upstream knowledge management"). New material includes:

• chapters on thesauri, controlled vocabularies, and metadata;
• new case studies;
• new print and electronic resources;
• a discussion of the practice of "information architecture."

Highlights
Some of my favorite sections are:

• "Metadata matrix" — a table that lists 16 types of metadata (from subject to document type) at 3Com. For each type, the authors provide a description, example, and estimate of how difficult the metadata is to maintain. For example, the "subject" metadata category is difficult to keep up to date, while the "geographic region" metadata category is easy.
• Fast and slow "layers" — two figures that compare the rate of change for societal segments and information architecture elements. For example, fashion and art change rapidly, but nature and culture change slowly. Content and user interfaces change rapidly, but classification schemes change slowly.
• Measuring value — specific examples of "opportunity cost" calculations using real companies and real numbers [this method computes the value of employee time consumed in searching for information]. Also presents the case against ROI calculations and offers other budget-justification approaches, such as story telling, education, competitive analysis, and others.
• Services — List and description of over 30 services that could be offered through an enterprise information architecture unit, including content acquisition, indexing, software evaluation, training, and liaison.
• Resources — The list of resources is extensive and includes both URLs and annotations.

Excerpts are available on the Web.

3. Enterprise Content Services
(Greg Laugero and Alden Globe, Addison-Wesley, 2002)
This book is also the successor to an earlier work, Managing Knowledge, written by the same authors and published in 1999. Like its predecessor, Enterprise Content Services analyzes Web content development through a corporate communications lens. Both books grew out of the authors' experiences as technical publications manager (Laugero) and webmaster (Globe) at the IT consulting firm J.D. Edwards (see our 1999 article on the J. D. Edwards "Knowledge Garden"). Both have a liberal arts education (philosophy and law for Globe, English and history for Laugero).

The first book was noteworthy for its:

• brevity (114 pages, lots of white space);
• simplicity (minimal jargon, good diagrams, lucid style);
• interdisciplinary focus (stress on a team approach that includes the author, "knowledge analyst", technology support consultant, "interface design director," and multimedia producer);
• bottom line emphasis (work flow diagrams that include time line, necessary content, staff involvement, and major milestones).

The second book is an expansion and update of the earlier material. The focus has shifted from how to use basic Web publishing technology to what the authors call content services— "a disciplined program for blending people, technology, processes, and content into a complete working solution for any organization." The first two chapters deal with strategic issues — the relationship of content to corporate objectives, how it adds value, how ROI can be measured. The remaining chapters deal with the nuts and bolts of managing content as a corporate asset. This section is divided into three parts:

1. Knowledge "story board" (business process analysis)

2. Content inventory

3. Enterprise content catalog

When we were asked to review the galley proofs, we noted that in Chapter 6, "The enterprise content catalog," there was no mention of librarians. In response, the authors included the following statement:

Highlights
There is little really new material in Enterprise Content Services. The highlights below are drawn from both books.

• lifecycle diagrams ("story boards") — a visual tool to help analyze information requirements. Diagrams in the second book are more detailed with a supporting "content inventory" table.

click image to enlarge

Lifecycle diagrams are a good way to visualize the users, content, and processes involved in a business process.

• action plan — a 90-day plan for improving a single business cycle. The plan includes a knowledge audit, selecting a team and a technology, marketing, and content maintenance.
• staff roles — roles and skills needed to produce quality content include manager, departmental knowledge steward, and individual authors. Specific jobs include writers, communications professionals, Web developers, HR professionals, IT staff, graphic artists, PR professionals, and competitive intelligence professionals.

A sample chapter is available for Managing Knowledge.

4. Metadata solutions
(Adrienne Tannenbaum, Addison-Wesley, 2002)
This is an industrial-strength introduction to metadata management written by a Rutgers M.B.A. with experience in database and application development. As head of her own consulting firm, she helps companies implement metadata gateways, Web-based data access, data integration for e-commerce, and other database tasks. (If you're not sure why metadata is important, see "Why is metadata a hot topic?")

This is a book written by an IT professional primarily for other IT staff — information systems managers, data managers, technical analysts and software developers. Although the writing style is relatively accessible to ordinary business people, the concepts can be very abstract (e.g. "meta-metamodels"), and the vocabulary is IT-oriented. The choice of "metadata solutions" as the title is significant. The author defines it as:

The key words here are "metadata," "integrated" and "logically connected." The main value of this book lies in a detailed discussion of five different metadata "solutions" to the problem of integrating specialized "islands of information" developed for specific applications. The five solutions are:

1. Centralized metadata repository — a "one-stop-shop" where metadata is consolidated from many data sources to facilitate metadata management and deployment. (For more on databases, the primary tool for creating repositories, see "Designing with databases.")

2. Integrated architecture — Metadata extracted from various applications is periodically loaded into a metadata repository in batch mode. Application program interfaces (API's) are used to allow various software applications to access metadata in the repository. (For an introduction to computer architecture, see "What is architecture?")

3. Information directory — a "yellow pages" list of information resources by category. The focus is on finding specific resources, not necessarily eliminating duplicate metadata or reusing metadata.

4. Metadata interchange — a system that involves the exchange of metadata from disparate sources, often through XML. Exchange is made possible through the use of standard tags (labels) and values.

5. Standalone metadata store — a "passive" database used to store metadata (i.e. codes, keywords, categories) for reference purposes. There is no metadata exchange or connectivity to other applications.

Each solution includes diagrams, examples, and references to metadata tools. To fully understand the solutions, you should read the earlier chapters, which are organized into four sections:

• Part I: Business context and overview of common (inadequate) solutions, such as Web directories and data warehouses
• Part II: Metadata overview — what it is, who benefits from it, how to identify it, how it's stored and accessed
• Part III: Metadata repositories and other metadata-based technologies
• Part IV: How to plan and implement a metadata solution

The solutions in Part V illustrate how it's possible to implement a full-blown metadata repository in phases, starting from a simple list of codes or categories in a spreadsheet program, then gradually moving toward a complex system with maintenance capabilities and real-time metadata access for both humans and computer programs.

Using a home-building analogy, this book tells you how to create the blueprint, select the materials, frame the structure, and add subsystems for plumbing, electricity, etc. It won't tell you how to decorate the home for a personal "look and feel." In other words, Metadata Solutions deals with technical structure, while Information Architecture deals with intellectual structure.

Highlights
Sections we found particularly interesting were:

• disasters in the making — explanation of why conventional solutions to metadata management don't work.
• 5 key questions that form a context for designing a metadata solution — what data do we have, what does it mean, where is it, how did it get there, how do I get it?
• metadata beneficiaries — Several charts show what kind of metadata is typically used by different staff roles, where it comes from, and whether it's common to all applications or specific to one application.
• non-metadata factors that influence solution development — e.g. "organizational readiness," the technical environment, and how to provide technical support.

A sample chapter is available.

5. Google Hacks
(Tara Calishain and Rael Dornfest, O'Reilly, 2003)
What does it mean when a book full of geeky tips, tricks, and program code shows up on the New York Times list of Paperback Business Best Sellers? Maybe a lot of people want to learn how to use their favorite search more effectively, get a higher ranking in Google's search results, or integrate custom searches into another application. This book has something for everyone, from beginners to experts.

Calishain is the editor of ResearchBuzz and a columnist for LLRX.com and Searcher magazine. Dornfest is a "maven," author, and conference organizer for O'Reilly, the publisher. The authors cover three main areas:

1. searching tips and tricks (for researchers and "power users");

2. how to interface Google with other programs using the Google API (for programmers and systems analysts);

3. legal ways to improve Web site visibility in Google (for webmasters and marketers).

Each of the 100 "hacks" in the book has an icon showing its level of difficulty — beginner, moderate, expert. The "expert" category is for people with some programming experience. The "beginner" category is for people with a basic knowledge of computers and the Internet. For Knowledge Base Publishers, the book's primary value is a greater understanding of how search engines work and how they can be interfaced with other information services.

What's a "hack?"
According to the authors, a "hack" is a quick-and-dirty solution to a programming problem or an interesting technique for getting a job done. Instead of providing an exhaustive manual on Google syntax or a dissertation on the computer science concepts used, the book shows techniques for getting the most out of the search engine.

What's an "API"?
API stands for "application program interface" — a software method of interacting with a program or the operating system in order to exchange data or initiate an action (e.g. update data). API's allow programmers to write routines that interact with commercial programs (without requiring access to proprietary code). They've become indispensable tools for interfacing systems and integrating information from different sources. (For additional information on integration and API's see "Content integration the easy way.")

What's the Google API?
The Google API is an application program interface that allows programmers to access the Google index (master list of Web pages) via a computer program. You can use it for personal, non-commercial use. To run the API, you need a set of computer instructions written in a programming language, such as the PERL, PHP, Java, or Python (the book gives examples in each language). Some of the things you can do with the API are:

• search Google on a specific word or phrase, then convert the results to a spreadsheet or database import format;
• add a "Google box" of search results to any Web page (the list can be updated automatically at a specific time of day);
• display a list of books along with referrals to them in Google and a list of related titles (uses both the Google and the Amazon.com API's).

Highlights
Here are some hacks we thought particularly interesting:

• Building Google Directory links using categories downloaded from the Open Directory Project. The Google Directory available to the public is very general. This hack lets you create a specialized directory by matching ODP categories with Google search results.
• Scraping Google results into a comma-delimited file (CSV). "Screen scraping" is a technique for importing search results into a format suitable for database or spreadsheet importing. This hack automates the screen scraping process of doing a Google search, savings the results page, and converting it into CSV format.
• Google interface for translators hack creates a customized search form for language translation — finding glossaries, locating parallel texts (identical pages in two or more languages), finding abbreviations. The technique can be used to find information in any specialized vocabulary.
• How to find reference sources — e.g. back issues of magazine and newspaper articles, directories, link lists, and other collections of information, technical definitions.
• Search engine optimization template — a Google-friendly HTML style sheet specification. A related hack lists 26 content management tips for generating Web site traffic. This list should be required reading for every Knowledge Base Publisher. The bottom line, according to the authors, is:

A description, including a table of contents and list of all the hacks, is available on the Web.

Summary
These five books offer Knowledge Base Publishers helpful tools, techniques, and concepts from three disciplines: knowledge management, library science, and information technology. There's something here for beginners and experts, neophytes to geeks. The following table summarizes their features.