基于多属性加权的社会化问答社区关键词提取方法
|
摘要
[目的 /意义]现有的关键词提取方法不适应社会化问答社区文本长度较短、内容表述口语化、数据集稀疏的特点,且很少考虑用户关注程度对词语重要性的影响,不能有效地提取此类文本的关键词,因此,提出针对社会化问答社区的多属性加权关键词提取方法。[方法 /过程]多属性加权关键词提取方法通过引入调节函数和词性对传统TF-IDF进行改进,并通过线性加权融合用户回答数、关注数、浏览数以及评论数4个用户关注属性来综合度量词语权重。[结果/结论]实验表明,该方法能更有效地提取社会化问答社区文本的关键词。
[Purpose/significance] Existing methods of extracting keywords can't be applied to the social Q&A community effectively,because they are not suitable for the characteristics of the social Q&A community which embodies short texts,colloquial contents and sparse data. They rarely think about the impact of users' attention on words. In view of the aforementioned problem,this paper presents a novel keywords extraction method based on multi-attributes weighted for the social Q&A community. [Method/process] This method improved the traditional TF-IDF algorithm by introducing the tuning function and the part of speech. Besides,it calculated the weight of words based on a linear weighting formula,which fused four attributes of user focus by dealing with numbers of users' answer,attention,browse,and comments. [Result/conclusion]Experiments show that this method can extract keywords from the social Q&A community more effectively.
[Purpose/significance] Existing methods of extracting keywords can't be applied to the social Q&A community effectively,because they are not suitable for the characteristics of the social Q&A community which embodies short texts,colloquial contents and sparse data. They rarely think about the impact of users' attention on words. In view of the aforementioned problem,this paper presents a novel keywords extraction method based on multi-attributes weighted for the social Q&A community. [Method/process] This method improved the traditional TF-IDF algorithm by introducing the tuning function and the part of speech. Besides,it calculated the weight of words based on a linear weighting formula,which fused four attributes of user focus by dealing with numbers of users' answer,attention,browse,and comments. [Result/conclusion]Experiments show that this method can extract keywords from the social Q&A community more effectively.
引文
[1]陈娟,邓胜利.社会化问答平台用户体验影响因素实证分析---以知乎为例[J].图书情报工作,2015,59(24):102-108.
[2]袁红,赵娟娟.问答社区中用户与资源互动研究[J].图书情报工作,2014,58(18):102-109.
[3]WITTEN I H,PAYNTER G W,FRANK E,et al.KEA:Practical automatic keyphrase extraction[C]//Proceedings of the fourth ACM conference on Digital libraries.New York:ACM,1999:254-255.
[4]HORITA K,KIMURA F,MAEDA A.Automatic keyword extraction for wikification of east asian language documents[J].International journal of computer theory and engineering,2016,8(1):32-35.
[5]方俊,郭雷,王晓东.基于语义的关键词提取方法[J].计算机科学,2008,35(6):148-151.
[6]费洪晓,康松林,朱小娟,等.基于词频统计的中文分词的研究[J].计算机工程与应用,2005,41(7):67-68.
[7]王立霞,淮晓永.基于语义的中文文本关键词提取方法[J].计算机工程,2012,38(1):1-4.
[8]黄鲁成,蒋林杉,苗红,等.基于网络问答社区的话题识别与分析---以知乎“老年人”话题为例[J].图书情报工作,2016,60(5):93-100.
[9]陈娟,高杉,邓胜利.社会化问答用户特征识别与行为动机分析---以“知乎”为例[J].情报科学,2017(5):69-74.
[10]傅柱,王曰芬,陈必坤.国内外知识流研究热点:基于词频的统计分析[J].图书馆学研究,2016(14):2-12.
[11]罗燕,赵书良,李晓超,等.基于词频统计的文本关键词提取方法[J].计算机应用,2016,36(3):718-725.
[12]陈伟鹤,刘云.基于词或词组长度和频数的短中文文本关键词提取方法[J].计算机科学,2016,43(12):50-57.
[13]张建娥.基于多特征融合的中文文本关键词提取方法[J].情报理论与实践,2013,36(10):105-108.
[14]张建娥.基于TFIDF和词语关联度的中文关键词提取方法[J].情报科学,2012(10):110-112,123.
[15]张瑾.基于改进TF-IDF方法的情报关键词提取方法[J].情报杂志,2014(4):153-155.
[16]廖晓,李志宏,席运江.基于加权知识网络的企业社区用户创新知识建模及分析方法[J].系统工程理论与实践,2016,36(1):94-105.
[17]SALTON G,BUCKLEY C.Term-weighting approaches in automatic text retrieval[J].Information processing&management,1988,24(5):513-523.
[18]PAIK J H.A novel TF-IDF weighting scheme for effective ranking[C]//Proceedings of the 36th international ACM SIGIR conference on research and development in information retrieval.New York:ACM,2013:343-352.
[19]施聪莺,徐朝军,杨晓江.TFIDF方法研究综述[J].计算机应用,2009,29(s1):167-170.
[20]罗繁明,杨海深.大数据时代基于统计特征的情报关键词提取方法[J].情报资料工作,2013,34(3):19-20.
[21]钱爱兵,江岚.基于改进TF-IDF的中文网页关键词抽取---以新闻网页为例[J].情报理论与实践,2008,31(6):147-152.
[22]张保富,施化吉,马素琴.基于TFIDF文本特征加权方法的改进研究[J].计算机应用与软件,2011,28(2):17-20.
[23]袁津生,毛新武.基于组合特征的中文新闻网页关键词提取方法[J].计算机工程与应用,2014,50(19):222-226.
[24]蒋昌金,彭宏,陈建超,等.基于主题词权重和句子特征的自动文摘[J].华南理工大学学报(自然科学版),2010,38(7):50-55.
[25]李湘东,巴志超,黄莉.一种基于加权LDA模型和多粒度的文本特征选择方法[J].现代图书情报技术,2015,31(5):42-49.
[26]路永和,王鸿滨.文本分类中受词性影响的特征权重计算方法[J].现代图书情报技术,2015,31(4):18-25.
[27]周鹏,蔡淑琴,石双元,等.基于关键词抽取的微博舆情事件内容聚合[J].情报杂志,2014(1):91-96.
[28]YE H M,CHENG W,DAI G Z.Design and implementation of on-line hot topic discovery model[J].Wuhan University journal of natural sciences,2006,11(1):21-26.
[29]SAATY T L.Modeling unstructured decision problems-the theory of analytical hierarchies[J].Mathematics and computers in simulation,1978,20(3):147-158.
[30]刘开第,庞彦军,周少玲,等.多准则排序中的路径问题及层次分析法推广[J].系统工程理论与实践,2015,35(4):973-983.
[31]邓爱东.多层次模糊综合评价法在图书馆危机管理中的应用[J].现代情报,2008,28(6):117-119.
[32]李亚平,焦建玲.网上交易流程效率评价[J].合肥工业大学学报:自然科学版,2009,32(8):1204-1207.
[2]袁红,赵娟娟.问答社区中用户与资源互动研究[J].图书情报工作,2014,58(18):102-109.
[3]WITTEN I H,PAYNTER G W,FRANK E,et al.KEA:Practical automatic keyphrase extraction[C]//Proceedings of the fourth ACM conference on Digital libraries.New York:ACM,1999:254-255.
[4]HORITA K,KIMURA F,MAEDA A.Automatic keyword extraction for wikification of east asian language documents[J].International journal of computer theory and engineering,2016,8(1):32-35.
[5]方俊,郭雷,王晓东.基于语义的关键词提取方法[J].计算机科学,2008,35(6):148-151.
[6]费洪晓,康松林,朱小娟,等.基于词频统计的中文分词的研究[J].计算机工程与应用,2005,41(7):67-68.
[7]王立霞,淮晓永.基于语义的中文文本关键词提取方法[J].计算机工程,2012,38(1):1-4.
[8]黄鲁成,蒋林杉,苗红,等.基于网络问答社区的话题识别与分析---以知乎“老年人”话题为例[J].图书情报工作,2016,60(5):93-100.
[9]陈娟,高杉,邓胜利.社会化问答用户特征识别与行为动机分析---以“知乎”为例[J].情报科学,2017(5):69-74.
[10]傅柱,王曰芬,陈必坤.国内外知识流研究热点:基于词频的统计分析[J].图书馆学研究,2016(14):2-12.
[11]罗燕,赵书良,李晓超,等.基于词频统计的文本关键词提取方法[J].计算机应用,2016,36(3):718-725.
[12]陈伟鹤,刘云.基于词或词组长度和频数的短中文文本关键词提取方法[J].计算机科学,2016,43(12):50-57.
[13]张建娥.基于多特征融合的中文文本关键词提取方法[J].情报理论与实践,2013,36(10):105-108.
[14]张建娥.基于TFIDF和词语关联度的中文关键词提取方法[J].情报科学,2012(10):110-112,123.
[15]张瑾.基于改进TF-IDF方法的情报关键词提取方法[J].情报杂志,2014(4):153-155.
[16]廖晓,李志宏,席运江.基于加权知识网络的企业社区用户创新知识建模及分析方法[J].系统工程理论与实践,2016,36(1):94-105.
[17]SALTON G,BUCKLEY C.Term-weighting approaches in automatic text retrieval[J].Information processing&management,1988,24(5):513-523.
[18]PAIK J H.A novel TF-IDF weighting scheme for effective ranking[C]//Proceedings of the 36th international ACM SIGIR conference on research and development in information retrieval.New York:ACM,2013:343-352.
[19]施聪莺,徐朝军,杨晓江.TFIDF方法研究综述[J].计算机应用,2009,29(s1):167-170.
[20]罗繁明,杨海深.大数据时代基于统计特征的情报关键词提取方法[J].情报资料工作,2013,34(3):19-20.
[21]钱爱兵,江岚.基于改进TF-IDF的中文网页关键词抽取---以新闻网页为例[J].情报理论与实践,2008,31(6):147-152.
[22]张保富,施化吉,马素琴.基于TFIDF文本特征加权方法的改进研究[J].计算机应用与软件,2011,28(2):17-20.
[23]袁津生,毛新武.基于组合特征的中文新闻网页关键词提取方法[J].计算机工程与应用,2014,50(19):222-226.
[24]蒋昌金,彭宏,陈建超,等.基于主题词权重和句子特征的自动文摘[J].华南理工大学学报(自然科学版),2010,38(7):50-55.
[25]李湘东,巴志超,黄莉.一种基于加权LDA模型和多粒度的文本特征选择方法[J].现代图书情报技术,2015,31(5):42-49.
[26]路永和,王鸿滨.文本分类中受词性影响的特征权重计算方法[J].现代图书情报技术,2015,31(4):18-25.
[27]周鹏,蔡淑琴,石双元,等.基于关键词抽取的微博舆情事件内容聚合[J].情报杂志,2014(1):91-96.
[28]YE H M,CHENG W,DAI G Z.Design and implementation of on-line hot topic discovery model[J].Wuhan University journal of natural sciences,2006,11(1):21-26.
[29]SAATY T L.Modeling unstructured decision problems-the theory of analytical hierarchies[J].Mathematics and computers in simulation,1978,20(3):147-158.
[30]刘开第,庞彦军,周少玲,等.多准则排序中的路径问题及层次分析法推广[J].系统工程理论与实践,2015,35(4):973-983.
[31]邓爱东.多层次模糊综合评价法在图书馆危机管理中的应用[J].现代情报,2008,28(6):117-119.
[32]李亚平,焦建玲.网上交易流程效率评价[J].合肥工业大学学报:自然科学版,2009,32(8):1204-1207.