Almost since the beginning, search engines have been an integral part of the Web experience. Given that history, you might think that they would be routine, well-established tools. But that is far from the case, according to two recent conferences.
The Association of Information and Dissemination Centers (ASIDIC) held its Spring 2005 Meeting March 21-23 in New Orleans. The theme was Search Wars and the Next Wave of Internet Innovation. Then, Infonortics, Ltd. held its 10th Annual Search Engine Meeting in Boston April 11-12. Although both conferences focused on search engines, the tone of each was quite different. The ASIDIC meeting centered more on the business and market aspects of search engines, while the Infonortics meeting emphasized new technologies and research advances.
ASIDIC featured two keynotes--one by Gary Price, editor of the Resource-Shelf Weblog and co-author of The Invisible Web, and the other by David Jastrow, senior analyst at Simba Information. Price believes that the information industry has done a poor job of keeping itself relevant in this one-stop-shopping age. He said that many users do not know about resources beyond the three major search engines (Google, Yahoo!, and MSN) and that the real invisible Web is everything beyond the first four or five search results. He identified some important trends including personalization, specialized search tools and engines, federated search, desktop searching, and multimedia searching. He also discussed the evolution of search engines into answer engines.
In his keynote, Jastrow touched on many of the same points as Price. He noted that as content becomes more complex, searchers want facts at their fingertips as well as more personalization and high-quality research. The ability to "buy by the drink" is particularly important, and new versions of some products (IEEE Xplore, for example) illustrate this fact. Jastrow's presentation also included market analyses for the legal, STM, and business areas. He pointed to the results of a survey that asked publishers if Google Scholar represented a threat or opportunity. More than half of the respondents thought that it would be an opportunity because it would bring more users to their Web sites. Librarians also viewed Google Scholar positively; more than half of the respondents said it would have a positive impact on their jobs. However, they did have one caveat: Google needs to improve its communications with publishers.
Jan Pedersen, chief scientist for Yahoo!, keynoted the Infonortics meeting. He noted that although online advertising is the fastest growing segment of the advertising business, a significant problem is matching ads appropriately to a user's search, showing that ontologies and keyword indexing continue to play an important role on the Web.
A few presentations on search models and techniques provided an interesting view of searching. Susan Feldman of IDC said that new search applications are hybrids of more complex technologies. Searching, she said, must be viewed in the context in which it is being used; the same rule base cannot be applied to all types of searches. She advised attendees to keep an eye on trends such as focusing search to the user's context without losing important information, providing uniform access to multiple document collections, and offering answers and guidance instead of lists of URLs. She also told us to watch for the diversification of search products.
Feldman also cited a recent IDC market study that revealed that knowledge workers spend up to 9.5 hours a week (and an enormous amount of industry dollars) on search- and information-related tasks.
According to Oshoma Momoh, general manager of MSN Search at Microsoft, search has become central to many other Web-related activities. Momoh's search immersion model (see Figure 1) described lessons Microsoft learned while developing MSN Search. …