Paper: SenseClusters - Finding Clusters That Represent Word Senses

ACL ID N04-3008
Title SenseClusters - Finding Clusters That Represent Word Senses
Venue Human Language Technologies
Session System Demonstration
Year 2004
Authors

SenseClusters is a freely available word sense discrimination system that takes a purely unsu- pervised clustering approach. It uses no knowl- edge other than what is available in a raw un- structured corpus, and clusters instances of a given target word based only on their mutual contextual similarities. It is a complete sys- tem that provides support for feature selec- tion from large corpora, several different con- text representation schemes, various clustering algorithms, and evaluation of the discovered clusters.