融合时序性和波动性的热点话题发现研究

doi:10.3969/j.issn.1000-1565.2018.04.012

河北大学学报(自然科学版) ›› 2018, Vol. 38 ›› Issue (4): 416-422.DOI: 10.3969/j.issn.1000-1565.2018.04.012

融合时序性和波动性的热点话题发现研究

李汉才¹,徐建民²,吴树芳^3,4

收稿日期:2017-10-12 出版日期:2018-07-25 发布日期:2018-07-25
通讯作者: 吴树芳(1979—),女,河北邯郸人,河北大学副教授,博士,主要从事信息检索、不确定信息处理方向研究.E-mail:shufang_44@126.com
作者简介:李汉才(1973—),男,河北枣强人,河北大学教授,主要从事信息检索方向研究. E-mail:hbdxlhc@163.com
基金资助:
国家社科基金资助项目(17BTQ068);河北省教育厅青年基金资助项目(QN2015099);河北大学中西部提升综合实力专项资金项目;河北省自然科学基金资助项目(F2015201142);中国博士后基金资助项目(2017M621078)

On hot topic detection by merging temporal and volatility

LI Hancai¹,XU Jianmin²,WU Shufang^{3, 4}

1.Personnel Department, Hebei University, Baoding 071002, China; 2.College of Computer Science and Technology, Hebei University, Baoding 071002, China; 3.College of Management and Economics, Tianjin University, Tianjin 300000, China; 4.College of Management, Hebei University, Baoding 071002, China

Received:2017-10-12 Online:2018-07-25 Published:2018-07-25

摘要/Abstract

摘要： 时序性和波动性直接与话题的热度有关,短时间内某话题出现的相关报道越多,则其热度越高;话题的波动幅度越大,则其热度越高.依据积分理论给出了基于时序性的相关报道密度计算和基于波动性的峰值计算,并采用线性调和的方法将二者融合,给出话题热度计算方法.实验采用TDT4语料作为测试集合,验证了该方法的有效性与合理性.

关键词: 热点话题, 发现, 时序性, 波动性

Abstract: Temporal and volatility are directly related to the hot degree of topic, namely, the more related stories the topic has in a certain time distance, and the larger fluctuation the topic is, the higher hot degree of this topic. Applying integration theory, methods of computing the density of related stories and the peak value are put forward based on temporal and volatility respectively. Finally, linear meditated method is used to merge them to compute the hot degree of a topic. Experiments are carried out on TDT4 corpus to testify the validity and rationality of our new method.

Key words: hot topic, detection, temporal, volatility

中图分类号:

TP390

李汉才,徐建民,吴树芳. 融合时序性和波动性的热点话题发现研究[J]. 河北大学学报(自然科学版), 2018, 38(4): 416-422.

LI Hancai,XU Jianmin,WU Shufang. On hot topic detection by merging temporal and volatility[J]. Journal of Hebei University (Natural Science Edition), 2018, 38(4): 416-422.

参考文献

[1] LI H, WEI J. Netnews bursty hot topic detection based on bursty features [Z]. International Conference on E-Business and E-Government, Washington DC, USA, 2010. DOI: 10.1109/ICEE.2010.365.
[2] XU G, QIU L, LIU H. Study on hot topic discovery from Chinese Texts [J]. International Journal of Information Management, 2014, 12(4): 267-273.
[3] 薛峰, 周亚东. 一种突发性热点话题在线发现与跟踪方法 [J]. 西安交通大学学报, 2011, 12(45): 64-70. DOI:10.7652/xjtuxb201112012. XUE F, ZHOU Y D. An online detection and tracking method for busty topics [J]. Journal of Xi’an Jiao Tong University, 2011, 12(45): 64-70. DOI:10.7652/xjtuxb 201112012.
[4] 陈羽中, 方明月, 郭文忠. 面向微博热点话题发现的多标签传播聚类方法研究 [J]. 模式识别与人工智能, 2015, 28(1): 1-10.DOI:10.16451/j.cnki.issn1003-6059.201501001. CHEN Y Z, FANG M Y, GUO W Z. Research on multi-lable propagation clustering method for microblog hot topic detection [J]. Pattern Recognition and Artificial Intelligence. 2015, 28(1): 1-10.DOI:10.16451/j.cnki.issn1003-6059.201501001.
[5] 魏德志, 陈福集, 林丽娜. 一种基于时间序列的热点话题发现模型和算法[J]. 情报科学. 情报科学, 2017, 10: 142-146. DOI: 10.13833/j.cnki.is.2017.10.026. WEI D Z, CHEN F J, LIN L N. A hot topic detection model and tracking algorithm based on time-series [J]. Information Science, 2017, 10: 142-146. DOI: 10.13833/j.cnki. is.2017.10.026.
[6] 孙明溪, 刘春琦. 基于DBSCAN算法与句间关系的热点话题发现研究[J].图书情报工作, 2017, 61(12): 113-121. DOI: 10.13266/j.issn.0252-3116.2017.12.015.
[7] 饶洁, 林育曼, 陈海媚. 基于粒子群算法的微博热点话题发现分析 [J]. 情报科学, 2016, 34(12): 51-54. DOI: 10.13833/j.cnki.is.2016.12.010. RAO J, LIN Y M, CHEN H M. Microblog hot topics tracking method based on particles swarm optimization [J]. Information Science, 2016, 34(12): 51-54. DOI: 10.13833/j.cnki.is.2016.12.010.
[8] 伍万坤, 吴清烈, 顾锦江. 基于EM-LDA综合模型的电商微博热点话题发现 [J]. 现代图书情报技术, 2015, 31(11): 33-40. DOI: 10.11925/infotech.1003-3513. 2015.11.06. WU W K, WU Q L, GU J J. Hot topic extraction from e-commerce microblog based on EM-LDA integrated model [J]. New Technology of Library and Information Service. 2015, 31(11): 33-40. DOI: 10.11925/infotech.1003-3513. 2015.11.06.
[9] 徐佳俊, 杨飏, 姚天昉,等. 基于LDA模型的论坛热点话题识别和追踪 [J]. 中文信息学报, 2016, 30(01): 43-49. XU J J, YANG Y, YAO T F, et al. LDA based hot topic detection and tracking for the forum [J]. Journal of Chinese Information Processing, 2016, 30(01): 43-49.
[10] 洪宇, 仓玉, 姚建民. 话题跟踪中静态和动态话题模型的核捕捉衰减[J]. 软件学报, 2012, 23(5):1100-1119. HONG Y, CANG Y, YAO J M. Descending kernel track of static and dynamic topic models in topic tracking [J]. Journal of Software,2012, 23(5):1100-1119.
[11] FUNG P C, YU J X, LIU H, et al. Time-dependent event hierarchy construction [Z]. The 13th ACM SIGKDD, New York, ACM, 2007. DOI: 10.1145/1281192.1281227.
[12] 叶雷, 高盛祥, 余正涛,等. 基于事件元素无向图的查询扩展方法 [J]. 中文信息学报, 2017, 31(1): 17-22. YE L, GAO S Y, YU Z T, et al. A query expansion method based on undirected graph of event elements [J]. Journal of Chinese Information Processing, 2017, 31(1): 17-22.
[13] ALLAN J, PAAPKA R, LAVRENKO V. On-line new event detection and tracking [Z]. The 21st Annual International ACM SIGIR Conference, New York: ACM, 1998.DOI: 10.1145/290941.290954.
[14] YANG Y, PIERCE T, CARBONELL J. A study on retrospective and on-line event detection [Z]. The 21st Annual International ACM SIGIR Conference, New York, ACM, 1998.
[15] CUTTING D R, KARGER D R, PEDERSEN J O, et al. Scatter/gather: a cluster-based approach to browsing large document collections [Z]. The 15th Annual International ACM SIGIR Conference, New York, ACM, 1992.
[16] 卢美莲, 王梓, 李佳珊. 利用改进LSH算法进行层次化新闻话题检测 [J]. 北京邮电大学学报, 2014, 37(3): 32-37. DOI: 10.13190/j.jbupt.2014.03.007. LU M L, WANG Z, LI J S. Hierarchical news topic detection using improved LSH [J]. Journal of Beijing University of Posts and Telecommunication, 2014, 37(3): 32-37. DOI: 10.13190/j.jbupt.2014.03.007.
[17] STEWART J. 微积分 [M]. 北京:高等教育出版社, 2014.
[18] 钱云. 非均衡数据分类算法若干应用研究 [D]. 长春: 吉林大学,2014. QIAN Y. Research on application of classification algorithm for imbalanced data [D]. Changchun: Jilin University,2014.
[19] GOUTTE C, GAUSSIER E. A probabilistic interpretation of precision, recall and F-score, with implication for evaluation [J]. International Journal of Radiation Biology, 2005, 51(5): 199-239. DOI: 10.1007/978-3-540-31865-1_25.

融合时序性和波动性的热点话题发现研究

On hot topic detection by merging temporal and volatility

PDF (PC)

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 7

编辑推荐

Metrics

本文评价

[1]	徐建民,李腾飞,吴树芳. 一种基于用户交互行为的微博社区发现方法[J]. 河北大学学报(自然科学版), 2016, 36(2): 189-196.
[2]	于红梅,张锋. 飞雁壮头蛛Chorizopes goosus的雄性新发现(蜘蛛目:园蛛科)[J]. 河北大学学报(自然科学版), 2014, 34(1): 70-72.
[3]	胡东升,张锋,李楠. 六盘盖蛛的雄性新发现(蜘蛛目:皿蛛科)[J]. 河北大学学报(自然科学版), 2011, 31(5): 528-531.
[4]	郭胜涛,张锋,王允. 米泉园蛛雄性新发现(蜘蛛目:园蛛科)[J]. 河北大学学报(自然科学版), 2011, 31(2): 192-194.
[5]	杨丽华,袁方,姚增利,王煜. 基于启发式规则的Deep Web接口发现[J]. 河北大学学报(自然科学版), 2010, 30(1): 107-112.
[6]	郭建宇,朱明生. 中国方胸蛛属1新纪录种及其雌性新发现(蜘蛛目:跳蛛科)[J]. 河北大学学报(自然科学版), 2010, 30(1): 93-96.
[7]	唐贵明,宋大祥,朱明生. 中国逍遥蛛属2新纪录种及1雄性新发现(蜘蛛目:逍遥蛛科)[J]. 河北大学学报(自然科学版), 2004, 24(4): 394-398.