An automatic method for generating sense tagged corpora
Rada F. Mihalcea, DAN I. MOLDOVAN · University of North Texas Digital Library (University of North Texas) · 1999
The unavailability of very large corpora with se-mantically disambiguated words is a major limi-tation in text processing research. For example, statistical methods for word sense disambiguation of free text are known to achieve high accuracy re-sults when large corpora are available to develop context rules, to train and test them. This paper presents a novel approach to automat-ically generate arbitrarily large corpora for word senses. The method is based on (1) the infor-mation provided in WordNet, used to formulate queries consisting of synonyms or definitions of word senses, and (2) the information gathered from Internet using existing search engines. The method was tested on 120 word senses and a pre-cision of 91 % was observed.