A probabilistic model of information retrieval: development and comparative experiments Part 2

The paper combines a comprehensive account of the probabilistic model of re trieval with new systematic experiments on TREC Programme material. It pres ents the model from its foundations through its logical development to cove r more aspects of retrieval data and a wider range of system functions. Eac h step in the argument is matched by comparative retrieval tests, to provid e a single coherent account of a major line of research. The experiments de monstrate, for a large test collection, that the probabilistic model is eff ective and robust, and that it responds appropriately, with major improveme nts in performance, to key features of retrieval situations. Part 1 covers the foundations and the model development for document collec tion and relevance data, along with the test apparatus. Part 2 covers the f urther development and elaboration of the model, with extensive testing, an d briefly considers other environment conditions and tasks, model training, concluding with comparisons with other approaches and an overall assessmen t. Data and results tables for both parts are given in Part 1. Key results are summarised in Part 2. (C) 2000 Elsevier Science Ltd. All rights reserved.