作者: Iryna Gurevych , Torsten Zesch
DOI:
关键词: Computer science 、 Natural language processing 、 Artificial intelligence 、 Semantic similarity
摘要: Wikipedia has been used as a knowledge source in many areas of natural language processing. As most studies only use certain snapshot, the influence Wikipedias massive growth on results is largely unknown. For first time, we perform an in-depth analysis this using semantic relatedness example application that tests wide range properties. We find almost no effect correlation measures with human judgments, while coverage steadily increases.