Retrieving terms and their variants in a lexicalized unification-based framework

Term extraction is a major concern for information retrieval. Terms are not fixed forms and their variations prevent them from being identified by a match with their initial string or inflection. We show that a local syntactic approach to this problem can give good results for both the quality of identification and parsing time. A specific tool, FASTR, is developed which handles an identification of basic terms and a parser of their variations as well. Terms are described by logic rules automatically generated from terms and their categonal structure. Variations are represented by metarules. The parser efficiently processes large size corpora with big dictionaries and mixes lexical identification with local syntactic analysis. We evaluate the accuracy of results produced by these metarules and improve these results with filtering metandes. 1 A Natural Language Processing Front-End to Automatic Indexing Complex terms represent important chunks of information in full-text documents and their identification plays a crucial role in information retrieval [1]: l Words in multiword terms are less ambiguous than in regular syntactic structures. An urtery can be either a road or a blood vessel while a Coronary artery only corresponds to the second meaning. l The entries of a thesaurus mainly consist of multiword terms. Their detection is an important clue fo