A PROBABILISTIC METHOD FOR COMPUTING TERM-BY-TERM RELATIONSHIPS

This article suggests a probabilistic method to compute the term relat ionships from relevance information, which complements the studies on a nonprobabilistic technique called pseudo-classification. A quadratic ranking function (i.e., a bilinear function) on the components of doc ument and query vectors is derived by incorporating the term-by-term r elationships. The conventional probabilistic indexing model, the proba bilistic retrieval model, and our earlier generalized model are specia l cases of the proposed model. By exploring the different views of pro bability, procedures for estimating the required parameters are provid ed.