Extracting Semantic Concept Relations from Wikipedia

Background knowledge as provided by repositories such as WordNet is of critical importance for linking or mapping ontologies and related tasks. Since current repositories are quite limited in their scope and currentness, we investigate how to automatically build up improved repositories by extracting semantic relations (e.g., is-a and part-of relations) from Wikipedia articles. Our approach uses a comprehensive set of semantic patterns, finite state machines and NLP-techniques to process Wikipedia de finitions and to identify semantic relations between concepts. Our approach is able to extract multiple relations from a single Wikipedia article. An evaluation for different domains shows the high quality and effectiveness of the proposed approach.