首页 | 本学科首页   官方微博 | 高级检索  
     检索      

基于“问题-用户”的网络问答社区专家发现方法研究
引用本文:龚凯乐,成颖.基于“问题-用户”的网络问答社区专家发现方法研究[J].图书情报工作,2016,60(24):115-121.
作者姓名:龚凯乐  成颖
作者单位:南京大学信息管理学院 南京 210023
基金项目:本文系国家自然科学基金面上项目“融合范式视角下的链接分析理论集成框架及其实证研究”(项目编号:71273125)和中国科学技术信息研究所合作研究项目“用户行为视角的链接动机与系统可用性研究”的研究成果之一。
摘    要:目的/意义] 以网络问答社区为研究对象,提出基于“问题-用户”传播网络的专家发现方法,为建立用户激励机制、完善专家推荐方法提供借鉴。方法/过程] 通过分析开放问答模式的特点,以“问题”和“用户”为节点、“答题关系”为有向边,构建“问题-用户”权威值传播网络,利用答案质量改进加权的HITS算法。结果/结论] 提出的算法可以较好地兼顾用户的答题数量与答案质量,能够选择出活跃度高、知识渊博的用户作为专家。

关 键 词:网络问答社区  专家发现  传播网络  链接分析  HITS算法  
收稿时间:2016-07-13

Finding Experts in Community Question Answering Based on “Question-User”
Gong Kaile,Cheng Ying.Finding Experts in Community Question Answering Based on “Question-User”[J].Library and Information Service,2016,60(24):115-121.
Authors:Gong Kaile  Cheng Ying
Institution:School of Information Management, Nanjing University, Nanjing 210023
Abstract:Purpose/significance] Taking Community Question Answering as the research object, this paper proposes an experts finding algorithm, which is based on "question-user" propagation network, to provide a reference for establishing incentive mechanism and improving experts recommendation.Method/process] By analyzing the features of open question answering model, this paper takes "question" and "user" as nodes, "answering question" as directed edges, and then builds an "question-user" authority propagation network. It also improves the weighted HITS algorithm by quality of answers.Result/conclusion] Experimental results show that the algorithm can achieve a balance between number and quality of users' answers, then select active and knowledgeable users as experts.
Keywords:Community Question Answering  experts finding  propagation network  link analysis  HITS algorithm  
点击此处可从《图书情报工作》浏览原始摘要信息
点击此处可从《图书情报工作》下载免费的PDF全文
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号