A study of probability kinematics in information retrieval

We analyze the kinematics of probabilistic term weights at retrieval time f or different Information Retrieval models. We present four models based on different notions of probabilistic retrieval. Two of these models are based on classical probability theory and can be considered as prototypes of mod els long in use in Information Retrieval, like the Vector Space Model and t he Probabilistic Model. The two other models are based on a logical techniq ue of evaluating the probability of a conditional called imaging; one is a generalization of the other. We analyze the transfer of probabilities occur ring in the term space at retrieval time for these four models, compare the ir retrieval performance using classical test collections, and discuss the results. We believe that our results provide useful suggestions on how to i mprove existing probabilistic models of Information Retrieval by taking int o consideration term-term similarity.