Automatic word sense discrimination

摘要: This paper presents context-group discrimination, a disambiguation algorithm based on clustering. Senses are interpreted as groups (or clusters) of similar contexts the ambiguous word. Words, contexts, and senses represented in Word Space, high-dimensional, real-valued space which closeness corresponds to semantic similarity. Similarity Space is second-order co-occurrence: two tokens contexts) word assigned same sense cluster if words they co-occur with turn occur training corpus. The automatic unsupervised both application: induced from corpus without labeled instances or other external knowledge sources. demonstrates good performance discrimination for sample natural artificial words.