Acquisition of causal information from medical documents in French for experts

This paper discusses how to select causal information from medical documents, in particular the information structured by verbal locutions in French. Causal relations are represented not only by verbs, X CAUSES Y, but also by other different information, for instance a temporal order between cause and effect, different meanings of the effect, and so on. With an example, we propose how to categorize causal relations in order to obtain much more detailed information from medical documents. This method is useful to experts to increase knowledge data base, and it could help students in Medicine to understand and organize their tasks.