摘 要:社区问答系统作为一种新兴的知识分享平台,在帮助用户获取有用信息方面取得了相当大的成功。将用户提出的问题推荐给感兴趣的回答者依然是社区问答平台面临的一个问题。针对兴趣度,已提出了多个表示模型,但这些模型没有考虑兴趣的时间维度。本文提出用TOT 主题模型建立备选回答者兴趣的动态变化模型,找出备选回答者的当前兴趣,然后进行问题推荐。实验表明本文提出的方法是有效的。
关键词:社区问答系统;问题推荐;兴趣度;Topics Over Time 主题模型
中图分类号:TP393.092;TP391.3 文献标识码:A 文章编号:2096-4706(2019)11-0001-04
Question Routing in Community Question Answering Based on Current Interest
ZHAO Yongbiao,ZHANG Qilin,GU Qiong
(Computer School of Hubei University of Arts and Science,Xiangyang 441053,China)
Abstract:Community question answering (CQA) has succeeded significantly in accessing useful information as a popularknowledge sharing platform. Recommending users’questions to interested respondents is still a problem facing the communityQA platform. For interest degree,several representation models have been proposed,but these models do not consider the timedimension of interest. In this paper,a TOT topic model is proposed to establish a dynamic change model of the interest of candidaterespondents,find out the current interest of candidate respondents,and then recommend questions. Experiments show that theproposed method is effective.
Keywords:community question answering;question recommendation;interest estimation;TOT topic model
基金项目:国家语委“十三五”科研规划项目:基于主题模型的Web 可比语料在线挖掘研究(项目编号:YB135-22);国家语委“十三五”科研规划项目:北宋书法家米芾书法字库创建及其推广应用(项目编号:YB135-33)。
参考文献:
[1] 曹艳蓉. 基于中文社区的智能问答系统的设计与研究 [D].南京:南京邮电大学,2018.
[2] 张力. 社区问答系统中答案排序和问题检索算法研究与应用 [D]. 合肥:中国科学技术大学,2018
[3] 刘晓鸣. 社区问答系统中的专家发现方法研究 [D]. 大连:大连理工大学,2013.
[4] 刘渊杰. 社区问答系统最佳回答机制的研究 [D]. 上海:上海交通大学,2010.
[5] 路遥. 用户交互式问答系统中问题推荐机制的研究 [D].合肥:中国科学技术大学,2012.
[6] 林鸿飞,王健,熊大平,等. 基于类别参与度的社区问答专家发现方法 [J]. 计算机工程与设计,2014,35(1):333-338.
[7] 戴秋敏. 互动问答平台专家发现及问题推荐机制的研究 [D]. 上海:华东师范大学,2014.
[8] 马泽锋. 基于机器学习的问答推荐系统问题推荐模型研究 [D]. 广州:中山大学,2013.
[9] Zhou G,Liu K,Zhao J. Joint Relevance and Answer QualityLearning for Question Routing in Community QA [C]. Hawaii:International Conference on Information and Knowledge Management,2012:1492-1496.
[10] 刘明荣. 协作式问答系统关键技术研究 [D]. 北京:中国科学院研究生院,2010.
[11] Li B,King I,Lyu M R. Question routing in communityquestion answering:Putting category in its place [C]. Glasgow:Proceedings of the 20th ACM Conference on Information and KnowledgeManagement,2011:2041-2044
[12] Zhou Y,Cong G,Cui B,et al. Routing Questions to theRight Users in Online Communities [C]. Data Engineering,2009. ICDE'09. IEEE 25th International Conference on. S.l.:s.n.,2009:700-711.
[13] Li B,King I. Routing questions to appropriate answerers inCommunity Question Answering services [C]. Toronto:Proceedings ofthe 19th ACM international conference on Information and knowledgemanagement,2010:1585-1588.
[14] Wang X,McCallum A.Topics over Time:A Non-MarkovContinuous-Time Model of Topical Trends [C]. Philadelphia:Proceedings of the 12th ACM SIGKDD international conference onKnowledge discovery and data mining,2006:44-54.
作者简介:赵永标(1980-),男,汉族,湖北洪湖人,讲师,硕士,研究方向:自然语言处理方面的教学与研究。