论文部分内容阅读
【目的】进行基于术语同义关系发现的知识组织系统整合研究。【方法】提出一种英文同义关系自动发现算法,涉及基于词形还原的词形归并以及基于同义关系传递和来源词表颗粒度控制的语义归并等综合方法。【结果】通过对多来源领域术语的大规模实验评估,并与已有整合知识组织系统进行多指标比较,获得较为满意的归并正确率,体现出良好的可行性及实用价值。【结论】本算法可应用于大规模领域知识组织系统的整合研究中,并对中文知识组织系统整合有一定借鉴意义。
【Objective】 The study of knowledge organization system integration based on the discovery of synonymous terms in terms of terms. 【Method】 An automatic algorithm for English synonymous relationship discovery is proposed, which involves the combination of word shape reduction based on word shape reduction and semantic combination based on synonymous relationship transmission and granularity control of source vocabulary. 【Result】 Through the large-scale experimental evaluation of terminology in multi-source fields and comparison with the existing integrated knowledge organization system for multi-index comparison, the satisfactory correct rate of incorporation was obtained, showing good feasibility and practical value. 【Conclusion】 This algorithm can be applied to the integration research of large-scale domain knowledge organization system and has certain reference meaning to Chinese knowledge organization system integration.