Description

Semantic relations have been for a long time subject of interest for linguists, philosophers, researchers in artificial intelligence and cognitive science. Theoretical linguists have collected, defined, and characterized semantic relations (synonymy, antonymy, hyponymy), which proved to be extremely useful for language engineers, helping them to design language resources in an economical way.

Semantic relations are fundamental for processing language, be it for analyzing or generating text, Information Extraction, Question Answering, knowledge acquisition, etc. As widely shown in the literature, semantic relations allow improving considerably the performance of NLP applications.

Next to the practical value, we would like to clarify at this workshop theoretical aspects of semantic relations and their potential heuristic value for other tasks than NLP.