International Journal of Applied Mathematics and Machine LearningAuthor's: Jianping Yu, Xiamei Yuan, Wenxue Hong, Shaoxiong Li and Deming Mei
Pages: [55] - [81]
Received Date: August 26, 2014
Submitted by: Jianqiang Gao.
In this article, the attribute partial order structure approach and fuzzy c-means cluster are applied into the solution of semantic merger of English modal verb in order to compare the performance of the two approaches. English modal verb should is chosen as the target word for semantic merger solution. First, two models for word sense disambiguation of English modal should are established by the two approaches, respectively. The accuracies of word sense disambiguation reach 91% for the attribute partial order structure approach and 90% for the fuzzy c-means cluster. Then, the semantic mergers of should are solved based on the two models of word sense disambiguation, respectively. Finally, the performance and accuracies by the two approaches are compared. The comparative result shows that, the fuzzy c-means cluster is easier in use than the attribute partial order structure approach, however, the attribute partial order structure approach performs much better than fuzzy c-means cluster in the solution of semantic merger of English modal verb.
semantic merger, solution, english modal verb, attribute partial order structure approach, fuzzy c-means cluster.
