A medical multilingual information retrieval

The Web is full of documents and resources. Users employ different strategies to find information they need: by browsing, using search engines, by following existing categories in a Web catalog. For technical sublanguages such as the medical one, document indexing based on lexical entities at a subword level has proved useful. However, it still remains challenging to identify and to delimit the meaningful lexical entities, as well as to group them in synonymy classes. We present a lexicographic and semantic foundation underlying the multilingual MORPHOSAURUS lexicon.