Mercedes, a Term-in-Context Highlighter
Proceedings of the Fourth International Conference on Language Resources and Evaluation (LREC 2004)
Abstract
It happens very often that researchers in Terminology need to know about the terms included in a given LSP corpus. One possibility is to run a term extractor but in this case such a tool provides just term candidates, but not valid terms. Therefore, it is mandatory a term validation process that is not always easy and affordable. A different option is to take profit of the public lexical resources (dictionaries and glossaries) that are increasingly freely available through the Internet. To accept only such a terms as the valid ones is usually enough for most of the researches. Mercedes, a term recogniser developed at the IULA, has been designed to fulfil this need. It integrates different public available lexical resources in order to provide researchers with a tool to show and highlight the valid terms (and its contexts) found in a certain text from a given domain. Mercedes consists of two main modules, the recogniser program and the dictionaries module. The recogniser can be run on specialised texts that have been previously tagged and lemmatised as part of the IULA's Corpus Tècnic. After running the recogniser, the user can then navigate the output with any web browser.