Publications

Presents a new model for probabilistic and document retrieval that uses logistic regression to calculate a best-fit matching equation for a database, given a set of potential retrieval clues.

Computer simulation of how different components of library service are interrelated.

Studies the use of Logistic regression methods to meld relevance clues into a single probability of relevance estimate. The model is applied to the domain of American case law.

A study of the optimizing of libraries' internal procedures.