Generating a multilingual taxonomy based on multilingual terminology clustering

• 来源期刊: 中国文献情报（英文刊）
• 关键词: Multilingual taxonomy multilingual terminology clustering cross-language terminology clustering parallel corpus mixed language
• 年，卷（期）: 2011,（2）
• 页码范围: 27-40
• 页数: 14页
• 语种: 英文

摘要

Taxonomy denotes the hierarchical structure of a knowledge organization system.It has important applications in knowledge navigation,semantic annotation and semantic search.It is a useful instrument to study the multilingual taxonomy generated automatically under the dynamic information environment in which massive amounts of information are processed and found.Multilingual taxonomy is the core component of the multilingual thesaurus or ontology.This paper presents two methods of bilingual generated taxonomy:Cross-language terminology clustering and mixed-language based terminology clustering.According to our experimental results of terminology clustering related to four specific subject domains,we found that if the parallel corpus is used to cluster multilingual terminologies,the method of using mixed-language based terminology clustering outperforms that of using the cross-language terminology clustering.