摘要 |
Disclosed is an improved word similarity calculation method. The method comprises: when one of a word A and a word B to be compared does not exist in a semantic dictionary, establishing an extension phrase a[M] of the word A and/or an extension phrase b[N] of the word B by means of a pre-set synonym dictionary; calculating a similarity value between the word A and each word in the extension phrase b[N], or a similarity value between each word in the extension phrase a[M] and the word B, or a similarity value between each word in the extension phrase a[M] and each word in the extension phrase b[N]; and taking a maximum similarity value as a similarity value between the word A and the word B. The embodiments of the present invention extend the synonyms of the word A and/or the word B to be compared, so that the accuracy for calculating the similarity between the words is improved, and the intelligence level of an intelligent interaction system is further improved. |
申请人 |
ANYCHECK INFORMATION TECHNOLOGIES CO.LTD.;E-TECHNO INFORMATION TECHNOLOGIES CO.,LTD.;BIO-TECH ACADEMY(CHINA) CO.,LTD. |
发明人 |
ZHANG, Guanjing;CHEN, Xingming;GE, Xinke;ZHANG, Shaopeng;FANG, Jingfang;GAO, Weiming;LIANG, Yanni;ZHOU, Rong;LIANG, Haoyuan;ZHOU, Liang |