您的位置:山东大学 -> 科技期刊社 -> 《山东大学学报(工学版)》

山东大学学报(工学版) ›› 2015, Vol. 45 ›› Issue (2): 37-42.doi: 10.6040/j.issn.1672-3961.2.2014.125

• 机器学习与数据挖掘 • 上一篇    下一篇

基于语义相似度的标签优化

钱肃驰, 彭甫镕, 陆建峰   

  1. 南京理工大学计算机科学与工程学院, 江苏 南京 210094
  • 收稿日期:2014-05-23 修回日期:2014-11-20 出版日期:2015-04-20 发布日期:2014-05-23
  • 通讯作者: 陆建峰(1969-),男,江苏淮安人,教授,博士,博士生导师,主要研究方向为数据挖掘,图像处理,智能机器人.E-mail:lujf@njust.edu.cn E-mail:lujf@njust.edu.cn
  • 作者简介:钱肃驰(1991-),男,江苏苏州人,硕士研究生,主要研究方向为数据挖掘.E-mail:qiansuchi2012@gmail.com
  • 基金资助:
    江苏省"六大人才高峰"资助项目

Tag optimization based on semantic similarity

QIAN Suchi, PENG Furong, LU Jianfeng   

  1. School of Computer Science and Engineering, Nanjing University of Science and Technology, Nanjing 210094, Jiangsu, China
  • Received:2014-05-23 Revised:2014-11-20 Online:2015-04-20 Published:2014-05-23

摘要: 为解决社交媒体中标签的缺失、错误等问题,提出一种基于内容相似度和语义相似度的标签优化方法。首先利用TF-IDF(term frequency—inverse document frequency)计算文本间相似度,然后利用文本间相似度与标签相似度的一致性定义了目标函数,最后加入了修正项来减少优化前后用户提供标签的偏差。将目标函数应用到豆瓣电影标签进行优化,并将结果与原标签进行比较分析。与原标签相比,优化后的标签准确性得到了提高。试验结果表明,该方法能够有效地优化标签,有效解决标签缺失和错误等问题。

关键词: 内容相似度, 标签优化, 语义相似度, 电影标签, 社交媒体

Abstract: To effectively solve those problems such as lack and misuse of tags in the social media, a tag optimization method based on content similarity and semantic similarity was proposed. Firstly, TF-IDF(term frequency—inverse document frequency) was used to calculate the text similarity. Afterwards, the objective function was defined by the consistency between text similarity and tag similarity. Finally, correction term was added in optimization process to reduce the deviation of tags provided by users. The objective function was applied to Douban Movie to optimize movie tags and the results were compared and analyzed with the original tags. The accuracy of the optimized tags was improved by comparison. Experimental results showed that the method could effectively optimize tags and solve those problems such as lack and misuse of tags.

Key words: semantic similarity, social media, content similarity, tag optimization, movie tags

中图分类号: 

  • TP391.3
[1] BOYD D M, ELLISON N B. Social network sites:definition, history, and scholarship[J]. Engineering Management Review, IEEE, 2010, 38(3):16-31.
[2] SIGURBJRNSSON B, VAN Zwol R. Flickr tag recommendation based on collective knowledge[C]//Proceedings of the 17th International Conference on World Wide Web. Beijing, China:ACM, 2008:327-336.
[3] 张斌,张引,高克宁,等. 融合关系与内容分析的社会标签推荐[J]. 软件学报, 2012, 23(3):476-488. ZHANG Bin, ZHANG Yin, GAO Kening,et al.Combining relation and content analysis for social tagging recommendation[J]. Journal of Software, 2012, 23(3):476-488.
[4] KRESTEL R, FANKHAUSER P, NEJDL W. Latent dirichlet allocation for tag recommendation[C]//Proceedings of the Third ACM Conference on Recommender Systems. New York, USA:ACM, 2009:61-68.
[5] SYMEONIDIS P, NANOPOULOS A, MANOLOPOULOS Y. Tag recommendations based on tensor dimensionality reduction[C]//Proceedings of the 2008 ACM Conference on Recommender Systems. Lausanne:ACM, 2008:43-50.
[6] JASCHKE R, MARINHO L, HOTHO A, et al. Tag recommendations in folksonomies[M]// Berlin Heidelberg:Springer, 2007:506-514.
[7] LEE SEUNG. Learning the parts of objects by non-negative matrix factorization[J]. Nature, 1999, 401(6755):788-791.
[8] CARBONELL J, GOLDSTEIN J. The use of MMR, diversity-based reranking for reordering documents and producing summaries[C]//Proceedings of the 21st Annual International ACM SIGIR Conference on Research and Development in Information Retrieval. New York, USA:ACM, 1998:335-336.
[9] 张雷鸣, 李秋丹, 廖胜才. 非负矩阵分解在标签语义分析中的应用[J]. 计算机科学, 2010, 37(4):171-174. ZHANG Leiming, LI Qiudan, LIAO Shengcai. Application of non-negative matrix factorization in tag semantics analysis[J]. Computer Science, 2010, 37(4):171-174.
[10] HOYER P O. Non-negative matrix factorization with sparseness constraints[J]. Journal of Machine Learning Research, 2004(5):1457-1469.
[11] JIANG B. Improving collaborative tag recommendation by using local lexicon in social comment context[C]//Proceedings of 15th International Conference on Computer Supported Cooperative Work in Design (CSCWD). Lausanne, Switzerland:IEEE, 2011:577-580.
[12] LIU D, HUA X S, WANG M, et al. Retagging social images based on visual and semantic consistency[C]//Proceedings of the 19th International Conference on World Wide Web. Raleigh, USA:ACM, 2010:1149-1150.
[13] HOFMANN T. Unsupervised learning by probabilistic latent semantic analysis[J]. Machine Learning, 2001, 42(1-2):177-196.
[14] BLEI D M, NG A Y, JORDAN M I. Latent dirichlet allocation[J]. Journal of Machine Learning Research, 2003(3):993-1022.
[15] MERIALDO B. Tagging English text with a probabilistic model[J]. Computational Linguistics, 1994, 20(2):155-171.
[16] 计智伟,胡珉,尹建新. 特征选择算法综述[J]. 电子设计工程,2011,19(9):46-51. JI Zhiwei, HU Min, YIN Jianxin. A survey of feature selection algorithm[J]. Electronic Design Engineering, 2011, 19(9):46-51.
[17] 夏天. 汉语词语语义相似度计算研究[J]. 计算机工程, 2007, 33(6):191-194. XIA Tian. Study on Chinese words semantic similarity computation[J]. Computer Engineering, 2007, 33(6):191-194.
[18] LIN D. Using syntactic dependency as local context to resolve word sense ambiguity[C]//Proceedings of the 35th Annual Meeting of the Association for Computational Linguistics and Eighth Conference of the European Chapter of the Association for Computational Linguistics. New York, USA: Association for Computational Linguistics, 1997:64-71.
[19] LIU D, WANG M, YANG Y, et al. Tag quality improvement for social images[C]//Proceedings of IEEE International Conference on Multimedia and Expo. New York, USA:IEEE, 2009:350-353.
[20] LEE D D, SEUNG H S. Algorithms for non-negative matrix factorization[J]. Advances in Neural Information Processing Systems, 2000(3):556-562.
[1] 林江豪,周咏梅,阳爱民,陈锦. 基于词向量的领域情感词典构建[J]. 山东大学学报(工学版), 2018, 48(3): 40-47.
[2] 徐庆, 段利国, 李爱萍, 阴桂梅. 基于实体词语义相似度的中文实体关系抽取[J]. 山东大学学报(工学版), 2015, 45(6): 7-15.
[3] 尹坤,尹红风*,杨燕,贾真. 基于SimRank的百度百科词条语义相似度计算[J]. 山东大学学报(工学版), 2014, 44(3): 29-35.
Viewed
Full text


Abstract

Cited

  Shared   
  Discussed   
No Suggested Reading articles found!