基于Word2vec的微博短文本分类研究

doi:10.3969/j.issn.1671-1122.2017.01.009

摘要/Abstract

摘要：

随着微博等社会化媒体的信息量急剧膨胀,人们迫切需要实现这些信息的自动分类处理,以帮助用户快速查找所需信息和过滤垃圾信息。针对传统文本分类模型存在的特征维数灾难、无语义特征等问题,文章基于Word2vec模型对微博短文本进行了分类研究。鉴于Word2vec模型无法区分文本中词汇的重要程度,进一步引入TFIDF对Word2vec词向量进行加权,实现加权的Word2vec分类模型。最后合并加权Word2vec和TFIDF两种模型,实验结果表明合并后模型分类准确率高于加权Word2vec模型和使用TFIDF的传统文本分类模型。

关键词: 短文本分类, Word2vec, TFIDF, 支持向量机

Abstract:

With the rapid expansion of new available information on Microblogging and other social media. Text automatic classification becomes imperative in order to help people locate the information he inquires and filter spam. Based on the characteristics of curse of dimensionality and lack of semantic features in Traditional text classification model, put forward a short text classify based on Word2vec model.Since Word2vec can not distinguish the weight of words, we applied weights using tf-idf weighting with Word2vec, implemented weighted Word2vec. Then we concatenated tf-idf with our word2vec weighted by tf-idf. Our results show that the combination of Word2vec weighted by tf-idf without stop words and tf-idf without stop words can outperform either Word2vec weighted by tf-idf without stop words and tf-idf with or without stop word.

Key words: short text classification, Word2vec, TFIDF, SVM

中图分类号:

TP309.2

张谦, 高章敏, 刘嘉勇. 基于Word2vec的微博短文本分类研究[J]. 信息网络安全, 2017, 17(1): 57-62.

Qian ZHANG, Zhangmin GAO, Jiayong LIU. Research of Weibo Short Text Classification Based on Word2vec[J]. Netinfo Security, 2017, 17(1): 57-62.

图/表 6

参考文献 33

[1]	刘丽清. 微博虽“微” 足值道尔——微博特性之浅析[J]. 东南传播, 2009 (11): 153-154.
[2]	崔争艳, CUIZheng-yan. 基于语义的微博短信息分类[J]. 现代计算机:专业版, 2010(8):18-20.
[3]	代六玲, 黄河燕, 陈肇雄. 中文文本分类中特征抽取方法的比较研究[J]. 中文信息学报, 2004, 18(1): 26-32.
[4]	JOACHIMS T.Text Categorization with Support Vector Machines: Learning with many Relevant Features[M]. Berlin Heidelberg: Springer, 1998.
[5]	SALTON G, WONG A, YANG C S.A Vector Space Model for Automatic Indexing[J]. Communications of the ACM, 1975, 18(11): 613-620.
[6]	BERRY M W.Survey of Text Mining[J]. Computing Reviews, 2004, 45(9): 548.
[7]	BANERJEE S, RAMANATHAN K, GUPTA A.Clustering Short Texts Using Wikipedia[C] // ACM. SIGIR 2007: Proceedings of the, International ACM SIGIR Conference on Research and Development in Information Retrieval, July 23 -27, 2007, Amsterdam, the Netherlands. New York: ACM, 2007:787-788.
[8]	HU X, SUN N, ZHANG C, et al.Exploiting Internal and External Semantics For the Clustering Of Short Texts Using World Knowledge[C] // ACM. ACM Conference on Information and Knowledge Management, CIKM 2009, November 2-6, 2009, Hong Kong, China. New York: ACM, November. 2009: 919-928.
[9]	LIU Z, YU W, CHEN W, et al.Short Text Feature Selection for Micro-Blog Mining[C] // IEEE. International Conference on Computational Intelligence and Software Engineering, December 10-12, 2010, New York. New York: IEEE, 2010: 1-4.
[10]	王盛,樊兴华,陈现麟. 利用上下位关系的中文短文本分类[J]. 计算机应用,2010,30(3): 603-606.
[11]	范云杰,刘怀亮. 基于维基百科的中文短文本分类研究[J]. 现代图书情报技术,2012(3): 47-52.
[12]	赵辉. 一种基于维基百科的中文短文本分类算法[J]. 图书情报工作,2013,57(11):120-124.
[13]	翟延冬,王康平,张东娜,等. 一种基于WordNet 的短文本语义相似性算法[J]. 电子学报,2012,40(3):617-620.
[14]	宁亚辉,樊兴华,吴渝. 基于领域词语本体的短文本分类[J]. 计算机科学,2009,36(3):142-145.
[15]	BOLLEGALA D, MATSUO Y, ISHIZUKA M. Measuring Semantic Similarity between Words Using Web Search Engines[EB/OL]. , 2016-12-2.
[16]	SAHAMI M, HEILMAN T D.A Web-based Kernel Function for Measuring the Similarity of Short Text Snippets[C] //IEEE. International Conference on World Wide Web, WWW 2006, May 23-26, 2006, Edinburgh, Scotland, UK. NewYork: IEEE, 2006: 377-386.
[17]	王鹏, 樊兴华. 中文文本分类中利用依存关系的实验研究[J]. 计算机工程与应用, 2010, 46(3): 131-133.
[18]	ZELIKOVITZ S, HIRSHi H.Improving Short Text Classification Using Unlabeled Background Knowledge to Assess Document Similarity[C] // IEEE. Proceedings of the seventeenth international conference on machine learning, June 29 - July 2, 2000, San Francisco, USA. San Francisco: Morgan Kaufmann Publishers Inc, 2000: 1183-1190.
[19]	BLEI D M, NG A Y, JORDAN M I.Latent Dirichlet Allocation[J]. the Journal of Machine Learning Research, 2003( 3): 993-1022.
[20]	PHAN X H, NGUYEN L M, HORIGUCHI S.Learning to Classify Short and Sparse Text & Web with Hidden Topics from Large-scale Data Collections[C] //北京航空航天大学,国际万维网会议委员会. The 17th International Conference on World Wide Web,April 21-25, 2008, Beijing, China. Beijing:国际万维网会议委员会,2008:91-100.
[21]	CHEN M, JIN X, SHEN D. Short Text Classification Improved by Learning Multi-granularity Topics[C] // International Joint Conference on Artificial Intelligence, July 16-22, 2011, Barcelona, Catalonia,Spain.
	New York: AAAI Press, 2011:1776-1781.
[22]	方东昊. 基于LDA 的微博短文本分类技术的研究与实现[D].沈阳:东北大学, 2011.
[23]	吕超镇,姬东鸿,吴飞飞. 基于LDA 特征扩展的短文本分类[J]. Computer Engineering and Applications, 2015, 51(4): 6-7.
[24]	MIKOLOV T, CHEN K, CORRADO G, et al.Efficient Estimation of Word Representations in Vector Space[J]. Computer Science, 2013(1):28-36.
[25]	LILLEBERG J, ZHU Y, ZHANG Y.Support Vector Machines and Word2vec for Text Classification with Semantic Features[C]// IEEE, International Conference on Cognitive Informatics & Cognitive Computing, July 6-8, 2015, Beijing, China. New York: IEEE, 2015:136-140.
[26]	李静梅, 孙丽华. 一种文本处理中的朴素贝叶斯分类器[J]. 哈尔滨工程大学学报, 2003, 24(1): 71-74.
[27]	张士豪,顾益军,张俊豪. 基于用户聚类的热门微博分类研究[J]. 信息网络安全,2015(7):84-89.
[28]	施聪莺, 徐朝军, 杨晓江. TFIDF 算法研究综述[J]. 计算机应用, 2009, 29(B06): 167-170.
[29]	张玉芳, 彭时名, 吕佳. 基于文本分类 TFIDF 方法的改进与应用[J]. 计算机工程, 2006, 32(19): 76-78.
[30]	张越今,丁丁. 敏感话题发现中的增量型文本聚类模型[J]. 信息网络安全,2015(9):170-174.
[31]	PATEL B, SHAH D.Significance of Stop Word Elimination in Meta Search Engine[C]//IEEE. International Conference on Intelligent Systems and Signal Processing, March 1-2, 2013, G H Patel College of Engineering and Technology, Vallabh Vidyanagar, Gujarat, India. New York: IEEE, 2013:52-55.
[32]	BRAZDIL P B, SOARES C, DA COSTA J P. Ranking Learning Algorithms: Using IBL and Meta-learning on Accuracy and Time Results[J]. Machine Learning, 2003, 50(3): 251-277.
[33]	WOLF L, HANANI Y, BAR K, et al.Joint word2vec Networks for Bilingual Semantic Representations[J]. International Journal of Computational Linguistics and Applications, 2014, 5(1): 27-44.