Web 2.0 and Value-Added Indexing
By Marydee Ojala
Editor • ONLINE
When information professionals justify the cost of premium content databases and combat the notion that everything is available at no charge on the Internet, they frequently point to the value add of indexing. However, the companies that create our value-added indexing are moving away from human indexing and toward automated indexing. Web 2.0 applications are encouraging user-generated tagging, which is an uncontrolled type of indexing. Some premium content sources, such as Engineering Village, are incorporating customer-generated tags into their fee-based products. How do these trends
affect the validity of our justifications for premium content? How can information professionals today truly guarantee that they’ve found the best, most relevant, and most reliable information when dealing with value-added indexing?
Two companies, Elsevier and IFIS, published new editions of their thesauri in February. The 5th edition of the Ei Thesaurus is in print and contains 1,075 new engineering terms that were selected by human indexers. The 8th edition of the FSTA Thesaurus, available in print and electronic formats, contains 700 new terms in food science, based on frequency of occurrence. The latter relies on the authors’ use of words rather than an established taxonomy.
New ONLINE columnist Bill Badke, self-described as a champion of information literacy, points out in his InfoLitLand column that it’s a bad idea to assume that people learn to handle information well simply by handling information. I’d add that just because a controlled vocabulary is in use, that doesn’t equate to its optimal use by searchers. It’s also dangerous to assume that Web search works like premium content services search and that all premium content works the same. It’s particularly treacherous to rely on user commentary as unbiased, honest, and reliable.
Perusing librarian discussion lists proves to me that not everybody in the profession understands the differences between a database aggregator and a database producer. Nor is there always complete agreement on what happens to a database when its producer turns it over to an aggregator. Does that value-added indexing remain, or does the aggregator change it? Is that a good thing or a bad thing?
Thoughtful indexing and well-constructed taxonomies are a boon to those searching for specific nuggets of information. Text mining in massive collections of data to discover new relationships and new knowledge, as described in this issue by Pam Kiser and Kathryn Lavengood, requires very different techniques. Going beyond indexing, text mining uses complex algorithms to analyze information.
Is creating a value-added indexing system or a customized corporate taxonomy antithetical to Web 2.0 and Library 2.0 concepts? If applied rigidly, then yes—it probably is. Applied intelligently, these systems and taxonomies can embody the best of 2.0 principles. In fact, the roots of tagging go deep in traditional online. Many database editors used index terms from controlled vocabularies as descriptor terms, but they added newer, uncontrolled terms in an identifier field. Web 2.0 opens up the identifier field to searchers as well as indexers. Whether it’s controlled or not, value-added indexing only truly adds value if it’s artfully constructed in the editorial stage and intelligently employed by searchers.
is the editor of ONLINE. Comments? E-mail letters
to the editor to