Bayesian indexing: the next craze in search algorithms?

Bayes’ theorem is about updating assumptions. It provides a mathematical basis for a heuristic search algorithm. A system using such an algorithm could make a kind of “best guess” about what the query is likely to mean. Some disadvantages include the need for more individualized information, a possible tendency to focus on a limited set of works, and the potential for encouraging slopping searching. The literature on heuristic systems is already substantial, and is relevant to all of us in the library and information field.