Paper: Unsupervised Learning Of Word Sense Disambiguation Rules By Estimating An Optimum Iteration Number In The EM Algorithm

ACL ID W03-0406
Title Unsupervised Learning Of Word Sense Disambiguation Rules By Estimating An Optimum Iteration Number In The EM Algorithm
Venue International Conference on Computational Natural Language Learning
Session Main Conference
Year 2003
Authors

In this paper, we improve an unsuper- vised learning method using the Expectation- Maximization (EM) algorithm proposed by Nigam et al. for text classification problems in order to apply it to word sense disambigua- tion (WSD) problems. The improved method stops the EM algorithm at the optimum itera- tion number. To estimate that number, we pro- pose two methods. In experiments, we solved 50 noun WSD problems in the Japanese Dic- tionary Task in SENSEVAL2. The score of our method is a match for the best public score of this task. Furthermore, our methods were con- firmed to be effective also for verb WSD prob- lems.