We discuss the contents, structure and uses of WordNet, a large machine-tractable lexical database developed and maintained at Princeton. This semantic network contains information about tens of thousands of English lexicalized concepts. WordNet enables automatic systems to "understand" the meanings of words, to measure their semantic similarity, and to disambiguate words with multiple meanings. A powerful tool for Natural Language Processing, WordNet is used in a wide range of applications including information retrieval, question answering, machine translation and automatic reasoning and inference.