An Approach of Text Sentiment Analysis for Public Opinion Monitoring System

Autor:	Yujiu Yang, Min Zeng, Wenhuang Liu
Rok vydání:	2013
Předmět:	Information retrieval business.industry Polarity (physics) Computer science Microblogging Bigram Sentiment analysis Public opinion Class (biology) Government (linguistics) ComputingMethodologies_PATTERNRECOGNITION Noun Social media business
Zdroj:	Springer Proceedings in Complexity ISBN: 9781461468790 CSWS
DOI:	10.1007/978-1-4614-6880-6_11
Popis:	With the thriving of microblog, a huge amount of people get involved in online life. This leads the government to intensify supervision on online remarks, and opinion polarity is what they care most. But microblog opinions contain several specialties from hotel remarks; they are format-free, short, and most express only one polarity. In this paper, we target on finding an appropriate polarity recognition method for public opinion supervision system. In our method, we explore new feature extraction rules which extract phizs, emotional nouns, verbs, adjectives, and bigrams as representative features. Then, we apply SVM to classify these online opinions into positive and negative class. Based on the crawled real-world datasets, our method can respectively achieve an accuracy of 80.1% and 87.4% for microblog reviews and traditional hotel remarks, so the proposed method is appropriate and effective for public opinion supervision system.
Databáze:	OpenAIRE
Externí odkaz:	https://explore.openaire.eu/search/publication?articleId=doi_________::ed3d30ed075563da8d829644169a07c5 https://doi.org/10.1007/978-1-4614-6880-6_11 Zobrazit plný text záznamu