A CONTRASTIVE APPROACH TO TERM EXTRACTION: CASE-STUDY FOR THE INFORMATION RETRIEVAL DOMAIN USING BAWE CORPUS AS AN ALTERNATIVE COLLECTION

This work describes a case-study on contrastive term extraction from a popular tutorial on information retrieval by C.D. Manning, P. Raghavan and H. Schutze. The results on extraction of terms are evaluated with the help of a reference index of terms compiled by the authors themselves. Execution of the case-study was motivated by 2 factors. Firstly, this work is part of a major project on automatic creation of an ontology in the field of information retrieval. Secondly, the authors also study applicability of the British Academic Written English (BAWE) as an alternative collection necessary for a contrastive extraction of terms. BAWE is a well-balanced and sufficiently representative corpus containing high-quality academic texts on 35 academic disciplines. The aim of this work is to show that the use of a contrastive approach with the help of such a balanced and representative alternative collection as BAWE allows to quite effectively solve the problem of automatic term recognition.