Monday, February 14, 2005

The Search Goes On

Dr. Marcia Bates, a UCLA professor, designed a theoretical system that leaves the classic IR model in the dust, and rightly so. Bates, continues in the direction that Belkin began paving earlier in the research, and questioned by Bates and others, “Why cannot the system make it possible for the searcher to express the need directly as they would ordinarily, instead of in an artificial query representation for the system’s consumption?” (409)
This is the basis for the berrypicking model, which is simply the act or art of searching in a way closely related to the way in which real people search for real information, and in what Bates feels is a model that an IR system must design and imitate to accommodate a human search, rather than designing a system that is more sophisticated in itself, but more difficult or ineffective for a human to use. In essence, like HDTV, the sharper image is not the most sophisticated picture.
Bates understands the nature of the human search and describes it simply as a “bit-at-a-time” retrieval, like berrypicking. This technique is far different from the linearity of previous models in that it acknowledges the shifting needs of the searcher as the search progresses, the various sources that must be consulted throughout the search, the various methodologies that are employed by most searchers when looking for something, and even the change in the nature of the search. Bates distinguishes between browsing and berrypicking, and advances our appreciation for browsing, not as a waste of time, but as a productive part of searching, and one that has meaning for search engines. But, how can browsing be incorporated into an online search to retrieve meaningful “hits” when the searcher is not quite certain about the search itself? Well, it’s a matter of serendipitous discovery, and stumbling upon information that changes the nature of the search, and gives the curvy shape to the journey.
Of course, not everyone is capable nor willing to be so intuitive, nor so brave as to admit that one’s initial thesis stinks, and one should take a new and better path in complete contradiction to one’s original ideas. I have and will, but some (most?) would rather stick with heartless ideas and at least finish within a reasonable time. Like Castaneda, I would rather choose the path with heart. Even if it means I have a heap of notes after 15 years? Yup. W.B. did, but maybe for different reasons. His briefcase was much heavier than mine.
Bates incorporates her findings into practical design aspects, which in the last 15 years have actually been built into some of our best search engines, and some which are yet to come, and some which seem at best, impractical and useless.
In 1989, the idea of flipping through a book on screen was not a reality as far as I know, and when Amazon first announced its page-turning tool in 2003 or 2004, everyone was amazed. Was this thanks to Bates? I have to think that the universe thinks of things together like some giant world brain (thanks to H.G. Wells for that idea).

4.2 Bates, M. (1989). The design of browsing and berrypicking techniques for the online search interface. Online Review, 13(5), 407-424.

No comments: