CAJ | 학술논문

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。
상사도계산시자동문답영역리적중요내용。위료보증후선답안집중각답안능구비합리적배서,해결전통자동문답계통불능고효적종합평개상사도문제,제출이용종합지수법대관건사상사도、어의상사도등진행종합평개,득도종합상사도。병침대부분후선답안용여신식과다,불리우답안제취적정황,설계료쇠감상사도삼수,용래해결구자용여신식대답안제취적영향。실험결과표명,종합지수법적상사도산법능구유효적제고문답적정학솔。
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.