QQ客服 官方微博 反馈留言

	Collaborative Data Cleaning for Sentiment Classification with Noisy Training Corpus
	Wan, Xiaojun
	2011
英文摘要	Labeled review corpus is considered as a very valuable resource for the task of sentiment classification of product reviews. Fortunately, there are a large amount of product reviews on the Web, and each review is associated with a tag assigned by users to indicate its polarity orientation. We can download such reviews with tags and use them as training corpus for sentiment classification. However, users may assign the polarity tag arbitrarily and inaccurately, and some tags are not appropriate, which results in that the automatically constructed corpus contains many noises and the noisy instances will deteriorate the classification performance. In this paper, we propose the co-cleaning and tri-cleaning algorithms to collaboratively clean the corpus and thus improve the sentiment classification performance. The proposed algorithms use multiple classifiers to iteratively select and remove the most confidently noisy instances from the corpus. Experimental results verify the effectiveness of our proposed algorithms, and the tri-cleaning algorithm is most effective and promising.; http://gateway.webofknowledge.com/gateway/Gateway.cgi?GWVersion=2&SrcApp=PARTNER_APP&SrcAuth=LinksAMR&KeyUT=WOS:000312259200027&DestLinkType=FullRecord&DestApp=ALL_WOS&UsrCustomerID=8e1609b174ce4e31116a60747a720701 ; Computer Science, Artificial Intelligence; Computer Science, Information Systems; Computer Science, Theory & Methods; EI; CPCI-S(ISTP); 0
语种	英语
DOI标识	10.1007/978-3-642-20841-6-27
内容类型	其他
源URL	[http://ir.pku.edu.cn/handle/20.500.11897/321250]
专题	信息科学技术学院
推荐引用方式 GB/T 7714	Wan, Xiaojun. Collaborative Data Cleaning for Sentiment Classification with Noisy Training Corpus. 2011-01-01.

个性服务

查看访问统计

相关权益政策

暂无数据

收藏/分享

除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。