期刊名称:International Journal of Innovative Research in Computer and Communication Engineering
印刷版ISSN:2320-9798
电子版ISSN:2320-9801
出版年度:2015
卷号:3
期号:7
DOI:10.15680/ijircce.2015.03071127
出版社:S&S Publications
摘要:This paper presents a new idea for sentiment analysis in twitter, especially for the Indian users.Sentiment analysis is one of the best tool to measure sentiments of the users hidden behind their text. But it is possiblethat sentiments are not analysed successfully due to some barriers.Indeed, the task of automatic sentiment recognitionin online text becomes more difficult for all the aforementioned reasons like limited size of character i.e. 140, unlimitedspelling mistakes, slang words and different languages. In our research, the primary and underlying idea is that the factof knowing how people feel about certain topics can be considered as a classification task and removing the languagebarrier using Google Translator. Twitter is used for the collection of data corpus in multilingual. Collected raw datasetis transformed into standard language i.e. English, and used for polarity classification of data corpus. Hence, we presenthow sentiment analysis can assist different languages, by Google translator and experimental results on the NaiveBayes Classifier and Maximum Entropy classifier and comparison the results of these two. Experimental results showthat our proposed techniques are efficient and performs better than previously proposed methods. We worked withHinglish, however, the proposed technique can be used with any other language.
关键词:Google translator; data corpus; lexicon; SentiWordNet; sentiment analysis; twitter; Naïve Bayes;Classifier and Maximum Entropy Classifier.