2004) Detection of Domain Specific Terminology Using Corpora Comparison

Identifying terms in specialized corpora is a central task in terminological work (compilation of domain-specific dictionaries), but is labour-intensive, especially when the corpora are voluminous which is often the case nowadays. For the past decade, terminologists and specialized lexicographers have been able to rely on term-extraction tools to assist them in the selection of terms. However, most term-extractors focus on the identification of complex terms. Although complex terms (cellular telephone) are central to terminology processing, retrieval of uniterms (telephone) is still a major challenge. This paper evaluates the usefulness of a corpora comparison approach in order to find pinpoint corpus specific words in order to identify uniterms in the field of telecommunications