基于BTM和加权K-Means的微博话题发现

doi:10.16088/j.issn.1001-6600.2019.03.008

摘要/Abstract

摘要： 为适应微博数据的短文本、低词频、缺乏语义表达等特殊性,提高话题发现的准确性,利于用户从大量微博数据中获取有用信息,本文提出一种基于BTM和加权K-Means方法实现微博话题发现。首先,针对微博数据稀疏性的问题,采用BTM模型对微博中的短文本进行建模,获得话题词;然后针对传统K-Means算法本身的缺陷,提出加权K-Means算法实现微博话题发现;最后实验验证本文的方法,实验结果表明,BTM和加权K-Means方法解决了微博数据高维度和稀疏性的问题,提高了热点话题发现的准确性和有效性。

关键词: BTM模型, 加权K-Means, 微博数据, 话题发现

Abstract: In order to adapt to special features of microblogging data, such as short texts, low word frequency, and lack of semantic expression, improve accuracy of topic discovery, and help users obtain useful information, a method based on BTM and weighting K-Means is proposed to achieve topic discovery. Firstly, faced with the problem of data sparsity, the text model is built based on the BTM model to obtain the topic words. Secondly, aimed at defects of the traditional K-Means algorithm itself, the weighting K-Means algorithm is proposed to obtain microblogging topics. Finally, experiments are conducted to validate the method of this paper. The experimental results show that the BTM and weighting K-Means method can solve problems of high dimensionality and sparsity of microblogging data, and it improves the accuracy and effectiveness of topic discovery.

Key words: biterm topic model(BTM), weighting K-Means, microblogging data, topic discovery

中图分类号:

TP391

陈凤,蒙祖强. 基于BTM和加权K-Means的微博话题发现[J]. 广西师范大学学报（自然科学版）, 2019, 37(3): 71-78.

CHEN Feng,MENG Zuqiang. Topic Discovery in Microblog Based on BTM and Weighting K-Means[J]. Journal of Guangxi Normal University(Natural Science Edition), 2019, 37(3): 71-78.

参考文献

[1] BLEI D M,NG A Y,JORDAN M I.Latent Dirichlet allocation[J].Journal of Machine Learning Research,2003,3: 993-1022.
[2] 谢昊,江红.一种面向微博主题挖掘的改进LDA模型[J].华东师范大学学报(自然科学版),2013(6):93-101.DOI: 10.3969/j.issn.1000-5641.2013.06.011.
[3] LIU Quanchao,HUANG Heyan,FENG Chong.Micro-blog post topic drift detection based on LDA model[C]// Behavior and Social Computing: LNCS Volume 8178,2013:106-118.DOI:10.1007/978-3-319-04048-6_10.
[4] GE Gaofei,CHEN Liping,DU Junping.The research on topic detection of microblog based on TC-LDA[C]//2013 15th IEEE International Conference on Communication Technology.Piscataway NJ:IEEE Press,2013:722-727.DOI:10.1109/ICCT.2013.6820469.
[5] YAN Xiaohui,GUO Jiafeng,LAN Yanyan,et al.A biterm topic model for short texts[C]//Proceedings of the 22nd International Conference on World Wide Web.New York,NY:ACM Press,2013:1445-1456.DOI:10.1145/ 2488388.2488514.
[6] CHENG Xueqi,YAN Xianhui,LAN Yanyan,et al.BTM:topic modeling over short texts[J].IEEE Transactions on Knowledge and Data Engineering,2014,26(12):2928-2941.DOI:10.1109/TKDE.2014.2313872.
[7] 张佳明,王波,唐浩浩,等.基于Biterm主题模型的无监督微博情感倾向性分析[J].计算机工程,2015,41(7): 219-223,229.DOI:10.3969/j.issn.1000-3428.2015.07.042.
[8] LI Weijiang,FENG Yanming,LI Dongjun,et al.Micro-blog topic detection method based on BTM topic model and K-means clustering algorithm[J]. Automatic Control and Computer Sciences,2016,50(4):271-277.DOI:10.3103/ S0146411616040040.
[9] 王亚民,胡悦.基于BTM的微博舆情热点发现[J].情报杂志,2016,35(11):119-124,140.DOI:10.3969/j.issn.1002-1965.2016.11.022.
[10]HE Xingwei,XU Hua,LI Jia,et al.FastBTM:reducing the sampling time for biterm topic model[J]. Knowledge-Based Systems,2017,132:11-20.DOI:10.1016/j.knosys.2017.06.005.
[11]ZHANG Peng,LI Bicheng,YANG Ruipeng.Research on the topic evolution of microblog based on BTM-LPA[C]// Proceedings of the International Conference on Computer Science and Technology.Singapore:World Scientific,2017:860-875.DOI:10.1142/9789813146426_0098.
[12]刘少鹏,印鉴,欧阳佳,等.基于MB-HDP模型的微博主题挖掘[J].计算机学报,2015,38(7):1408-1419.DOI: 10.11897/SP.J.1016.2015.01408.
[13]黄发良,冯时,王大玲,等.基于多特征融合的微博主题情感挖掘[J].计算机学报,2017,40(4):872-888. DOI:10.11897/SP.J.1016.2017.00872.
[14]GEMAN S,GEMAN D.Stochastic relaxation, gibbs distributions and the Bayesian restoration of images[J]. Journal of Applied Statistics,1993,20(5/6):25-62.DOI:10.1080/02664769300000058.
[15]FENG Jun,FANG Yu.Research on hot topic discovery technology of micro-blog based on biterm topic model[C]//Geo-Spatial Knowledge and Intelligence: 4th International Conference on Geo-Informatics in Resource Management and Sustainable Ecosystem.Berlin:Springer,2016:234-244.DOI:10.1007/978-981- 10-3969-0_27.
[16]谢修娟,李香菊,莫凌飞.基于改进K-means算法的微博舆情分析研究[J].计算机工程与科学,2018,40(1):155-158.DOI:10.3969/j.issn.1007-130X.2018.01.023.
[17]ZHANG Huaping,YU Hongkui,XIONG Deyi,et al.HHMM-based Chinese lexical analyzer ICTCLAS[C]// Proceedings of the second SIGHAN workshop on Chinese language processin:Volume 17.Stroudsburg,PA: Association for Computational Linguistics,2003:184-187.DOI:10.3115/1119250.1119280.
[18]刘泽锦,王洁.同主题词短文本分类算法中BTM的应用与改进[J].计算机系统应用,2017,26(11):213-219.DOI: 10.15888/j.cnki.csa.006071.
[19]李卫疆,王真真,余正涛.基于BTM和K-means的微博话题检测[J].计算机科学,2017,44(2):257-261,274.DOI: 10.11896/j.issn.1002-137X.2017.02.042.

相关文章 15

[1]	张灿龙, 李燕茹, 李志欣, 王智文. 基于核相关滤波与特征融合的分块跟踪算法[J]. 广西师范大学学报（自然科学版）, 2020, 38(5): 12-23.
[2]	王健, 郑七凡, 李超, 石晶. 基于ENCODER_ATT机制的远程监督关系抽取[J]. 广西师范大学学报（自然科学版）, 2019, 37(4): 53-60.
[3]	肖逸群, 宋树祥, 夏海英. 基于多特征的快速行人检测方法及实现[J]. 广西师范大学学报（自然科学版）, 2019, 37(4): 61-67.
[4]	王勋, 李廷会, 潘骁, 田宇. 基于改进模糊C均值聚类与Otsu的图像分割方法[J]. 广西师范大学学报（自然科学版）, 2019, 37(4): 68-73.
[5]	张随远, 薛源海, 俞晓明, 刘悦, 程学旗. 多文档短摘要生成技术研究[J]. 广西师范大学学报（自然科学版）, 2019, 37(2): 60-74.
[6]	孙容海, 施林甫, 黄丽艳, 唐振军, 俞春强. 基于图像插值和参考矩阵的可逆信息隐藏算法[J]. 广西师范大学学报（自然科学版）, 2019, 37(2): 90-104.
[7]	朱勇建, 彭柯, 漆广文, 夏海英, 宋树祥. 基于机器视觉的太阳能网版缺陷检测[J]. 广西师范大学学报（自然科学版）, 2019, 37(2): 105-112.
[8]	王祺, 邱家辉, 阮彤, 高大启, 高炬. 基于循环胶囊网络的临床语义关系识别研究[J]. 广西师范大学学报（自然科学版）, 2019, 37(1): 80-88.
[9]	武文雅, 陈钰枫, 徐金安, 张玉洁. 基于高层语义注意力机制的中文实体关系抽取[J]. 广西师范大学学报（自然科学版）, 2019, 37(1): 32-41.
[10]	岳天驰, 张绍武, 杨亮, 林鸿飞, 于凯. 基于两阶段注意力机制的立场检测方法[J]. 广西师范大学学报（自然科学版）, 2019, 37(1): 42-49.
[11]	余传明, 李浩男, 安璐. 基于多任务深度学习的文本情感原因分析[J]. 广西师范大学学报（自然科学版）, 2019, 37(1): 50-61.
[12]	林原, 刘海峰, 林鸿飞, 许侃. 基于损失函数融合的组排序学习方法[J]. 广西师范大学学报（自然科学版）, 2019, 37(1): 62-70.
[13]	万福成,马宁,何向真. 融合事件特征及语义角色标注的藏文信息抽取技术[J]. 广西师范大学学报（自然科学版）, 2018, 36(2): 18-23.
[14]	夏海英,刘伟涛,朱勇建. 一种改进的快速SUSAN棋盘格角点检测算法[J]. 广西师范大学学报（自然科学版）, 2018, 36(1): 44-52.
[15]	梁晓萍,罗晓曙. 基于遗传自适应的维纳滤波图像去模糊算法[J]. 广西师范大学学报（自然科学版）, 2017, 35(4): 17-23.

Metrics

Viewed

Full text

Abstract

Cited

Shared

Discussed