Towards Intelligent Information Retrieval: An Overview of IR Research at U. Mass

The Smart information retrieval project was started in 1961. During the past 30 years methods have been developed for the automatic term assignment to natural-language texts (automatic indexing), automatic document clustering, collection searching, and the automatic reformulation of search queries using relevance feedback. Many of these procedures have been incorporated into prac tical retrieval settings. Although there is no hope of solving the content analysis problem for naturallanguage texts completely satisfactorily, the possibility of automatically analyz ing very large text samples offers new approaches for automatic text processing and information retrieval. Some methods for the massive analysis of natural language text are outlined together with applications in information retrieval. 1 The Vector Processing System Conventional information retrieval systems are based on Boolean query formu lations where keywords are used together with connecting Boolean operators. By constructing large so-called inverted indexes that contain for each allowable keyword the lists of addresses of all documents indexed by that keyword, it is possible to determine the set of documents corresponding to a given Boolean query formulation from the information stored directly in the index. This im plies that rapid responses can be provided in a conventional retrieval setting using standard Boolean processing methods. The conventional Boolean search system does, however suffer from a number of serious disadvantages: First, the Boolean logic remains inaccessible to many Department of Computer Science, Cornell University, Ithaca, NY 14853-7501. This study was supported in part by the National Science Foundation under grant 1ST 84-02735.