期刊名称：International Journal of Computer Science and Network Solutions
印刷版ISSN：2345-3397
出版年度：2015
卷号：3
期号：12
页码：1-12
出版社：International Journal of Computer Science and Network Solutions
摘要：In today's competitive business, paying attention to the feedback from customers has become a valuablefactor for organizations. Organizations have found that satisfied customers are not only a repeated buyer,they are also propaganda arm of the organization. Therefore, the correct analysis of their feedback byrelying on information technology tools is a key element in the success of the organizations in trade. Peoplegenerally share their opinions about purchased goods on the Web sites or in social networks. Extraction ofthese opinions is known as a special branch of text mining under the term of sentiment mining. Althoughthis category is brand new, but in recent years, extensive researches have been done on sentiment analysisand classification of intentions. Therefore, in this paper a model is suggested about sentiment mining withthe ability to extract users' opinion and product features. So dataset of customer comments has been madein a way that the comments are taken from a Website about some specific digital products. Then theparagraphed opinions are converted into sentences and the sentences are separated into two categories ofsubjective and objective. Next, user’s opinion and product features are taken from subjective sentences byusing StanfordPOStagger and relying on Tf-idf factor for product features and finding opinion polarity byusing SentiWordNet tools. In this way, user satisfaction of specific features of the product can be detected.As a means of evaluation, three factors of Recall, Precision and F-Measure provide an indication of theaccuracy of each part of this research
关键词：Opinion mining; Sentiment mining; Subjective and objective sentences; Text mining
