International Academic Journal of Science and Engineering

  • ISSN 2454-3896

Increase Business Intelligence Based on opinions mining in the Persian Reviews

Hamed Azizi Vamerzani and Maryam Khademi

Abstract: By improving electronic business and managing system of customer‟s request, a huge load of Textual data has produced direct and indirect by users. Customers of one specific product and even managers need to observe comprehensive and effective information of obtained opinions in order to make the best decision in minimum time. Research shows that 81% of internet users have searched on the website before buying a product or their accessories. Unfortunately many of these researches are allocated to English language and just few projects have been done in Persian. In this paper, the provided framework has the ability of predicting the polarity of a text by user‟s textual review and extracting the critical Features and then by means of an improved SVM model, user‟s opinion about a product and its Features has been classified. The trained model can predict the polarity with a high degree of accuracy

Keywords: Sentiment Analysis, Opinion Mining, Persian Language, Text Mining, Natural Language Processing

Page: 164-174

Volume 2, Issue 2, 2015