微博汽车领域中用户观点句识别方法的研究AStudyofMicroblog.PDFVIP

微博汽车领域中用户观点句识别方法的研究 潘艳茜，姚天昉 上海交通大学计算机科学与工程系，上海，200240 E-mail: pyx0218@ yao-tf@ 摘 要：该文主要研究如何自动识别微博中用户对各品牌汽车进行评价的句子。针对微博中汽车宣传信息 较多而由真正汽车用户发出的观点句所占比例很小的特点，该文提出了结合微博和汽车评论语料的基于 SVM 模型的分类方法。选取的特征包括词语、评价词个数、与评价对象有关的词语以及微博相关特征。 实验表明，评价词特征和部分微博相关特征可有效提高分类器性能，使用微博和汽车评论两种语料进行训 练的分类器性能要比仅使用微博语料的方法好。 关键词：微博；观点句识别；意见挖掘；SVM 中图分类号：TP391 文献标识码：A A Study of Microblog Customer Opinioned Sentences Recognition in Automobiles Domain Pan Yanxi, Yao Tianfang Department of Computer Science and Engineering, Shanghai Jiao Tong University, Shanghai 200240 E-mail: pyx0218@ yao-tf@ Abstract: This paper mainly investigated how to automatically recognize the customer opinion towards certain automobiles in microblogs. Since there are a lot of advertises and release information of cars in microblogs, customer-generated opinioned sentences are sparse. Therefore, this paper proposed a SVM classifier-based method that combines microblog data and car review data for training. The selected features included words, the number of opinion words, words that have certain relations with opinion targets, as well as microblog-related features such as emoticons and user type. Experiment results indicates that opinion words feature and some of the microblog-related features can effectively boost the performance of the classifier. In addition, the performance of the classifier that uses two kinds of data for training is better than the one that only uses microblog data. Keywords: Microblog ; Opinioned Sentences Recognition; Opinion Mining; SVM 1 引言 近年来，微博逐渐成为人们表达自己看法，交流意见的平台。微博中包含的大量主观 性信息在许多应用领域有很大的潜在价值。然而人工浏览这些信息成本高，效率低。所以， 针对微博文本的情感分析成为国内外的一个研究热点。与传统的在产品评论等领域的情感 分析相比，由于微博主题包罗万象且内容简短，对其进行情感分析更加困难。目前国外有 1 关英文Twitter 的情感分析研究已有较为成熟的成果，但基于中文微博的相关研究工作相对 较少。中文中经常会使用比喻、反讽等修辞手法