Friday, February 10, 2006

Teragram: Parsing the Web from the Inside Out

With its foundation in linguistic analysis, Teragram has been working behind the scenes since its founding in 1997 to become a leading provider of tools that help make searching more effective. Currently, Teragram provides a set of tools for automatic categorization, taxonomy management, entity extraction, and other text analysis tools that can be used to improve the accuracy and relevance of search. Note, Teragram also offers tools to enhance search effectiveness in 30 European, Middle-Eastern and Asian languages.

A recent deal with Reed Business Information (RBI) provides a hint of some things to come from Teragram. RBI will deploy Teragram's technology to categorize its large portfolio of business content, which will enable RBI sites to direct users to related content across the RBI universe. A potential added benefit of having its repository of content categorized into a common taxonomy would involve placing relevant ads alongside RBI content, provided that the ad placement technology leverages key words to determine relevance. Because the content tagging will be carried out by the publishers of the original content--conceivably by the same writers and editors who write the original content, as opposed to aggregators, data mining companies, or search engines--the quality of the tags assigned will be high.

The early web search engines and contextual ad networks mostly relied on guesstimates of the relevance of search results and ad placements based on a general analysis of the text on a page. We all know from experience that it can be difficult to find very targeted pieces of information using Google or Yahoo!, especially when searching a popular topic. However, with publishers of orginal content increasingly adding metadata that describe information down to a level of specificity such as the latest sales figures for a company or the date of birth of an executive, the accuracy and usefulness--and consequently the value--of the content increases enormously. We are beginning to see the results of investments by publishers in content management systems and taxonomy tools such as those offered by Teragram. As a result, finding information on the Web will become a much more satisfying experience in the coming years. Furthermore, applications that mine Web content, as well as specialized collections of premium content, should provide a much higher level of accuracy and relevance than we have seen on the Web thus far.
Post a Comment