基于CIM的相似度综合评价算法

The Comprehensive Index Method Evaluation in the Similarity

相似度计算是自动问答领域里的重要内容。为了保证候选答案集中各答案能具备合理的排序,解决传统自动问答系统不能高效的综合评价相似度问题,提出利用综合指数法对关键词相似度、语义相似度等进行综合评价,得到综合相似度。并针对部分候选答案冗余信息过多,不利于答案提取的情况,设计了衰减相似度参数,用来解决句子冗余信息对答案提取的影响。实验结果表明,综合指数法的相似度算法能够有效的提高问答的正确率。

The similarity calculation is an important part in the field of automatic question answering.In order to ensure the candidate answers can have a reasonable sort,solve the traditional automatic answering system is not efficient evaluation of similarity problem,comprehensive index method is proposed to carry out a comprehensive evaluation keywords similarity and semantic similarity,the result is integrated similarity.Against part of the candidate answers have too many redundant information and not conducive to the answer extraction,designed attenuation similarity parameters,used to address the impact of the sentence redundant information extracted from the answers.Experimental results show that composite index method of similarity algorithm can effectively improve the accuracy rate of the questions and answers.