基于关系因子似然推理的语义传递模型研究  

Research on Semantic Transitivity Model Based on Plausible Inference Using Relation Elements

作　　者：裘江南[1] 王带弟[1] 任利平[1] 

机构地区：[1]大连理工大学管理科学与工程学院,辽宁大连116024

出　　处：《情报科学》2015年第9期8-14,共7页Information Science

基　　金：国家自然科学基金(71073017)

摘　　要：现有的基于关系因子的似然推理能够从语义关系的本质特征出发获取语义传递规律,但其中关系因子的选取和取值依赖于人的主观判断和经验,而且结果关系类型的判断方法也有待进一步改进。针对这些问题,本文通过分析现有语义传递研究方法的特点和语义传递的影响因素,构建了基于关系因子似然推理的语义传递模型。根据现有客观知识体系中的语义关系可组合规律获取关系因子可组合矩阵,以此引入语境因素来对关系因子取值进行修正,从而提高其客观性。此外,还提出了基于向量空间模型的新型结果关系类型判断方法。最后在Word Net2.1中进行实验,对比分析实验结果和已有结果,验证了模型在语义传递判断方面的有效性。The existing plausible inference using relation elements can obtain semantic transitivity rulesfrom the intrinsic nature of semantic relations. However, the selection of relation elements and values as-signed for them are subjective and method for judging the resultant relation needs to be refined. Aiming atthese problems, the paper analyses the characteristics of the existing research methods and introduces theinfluencing factors of semantic transitivity, then generates the relation elements combination matrix ac-cording to the semantic relations combination rules, which can be used to revise the values of relation ele-ments to increase their objectivity. In addition, a novel method based on the vector space model for judg-ing the resultant relation is proposed. Finally, we conduct an experiment in Word Net2.1 and analyze theresult comparing with the existing result, which approves the validity of the model.

关 键 词：客观知识体系 语义传递 关系因子 似然推理 

分 类 号：G254[文化科学—图书馆学]