基于网络知识资源的术语相似度计算方法综述

Research on Term Similarity Computational Methods Based on Network Knowledge Resources

在对当前基于网络知识资源的术语相似度计算相关文献进行分析研究的基础上,将其归纳为基于知识资源的术语相似度计算思路、基于搜索引擎的术语相似度计算思路以及术语相似度指标集成计算思路。对各类计算思路的特点和尚存在的不足进行了总结。最后对基于网络知识资源的术语相似度计算的未来研究发展进行展望,旨在为进一步构建高效的术语相似度计算系统提供良好借鉴。

Based on the analysis of recent related literatures and projects, the paper concludes the term computational methods that based on network knowledge resources as follows： similarity computational methods based on knowledge resources, similarity computational methods based on search engine, and integration methods of multi-measure methods. It also concludes characters and shortcomings of those methods. Finally, the future development of research on term similarity measure based on network knowledge resources is discussed to help building more efficient term similarity calculation system in further step.