Rating Prediction of Social Sentiment from Textual Review by Recognizing Contextual Polarity

Improving Recommendation Accuracy and User Perception with Sentiment Analysis

Abstract

In recent years, we can see various website on user can provide hisher reviews for product they have purchased. However mining valuable information from these reviews for recommendation of product crucial task. Traditionally for recommendation of product various factor are considered like user purchase record, uses location, product category etc. In our system we are proposing the sentiment-based rating prediction method which will improve the recommendation prediction accuracy. This system uses dictionary based classification for accurately classifying the reviews as positive, negative and neutral. In this system social user reviews goes through POS tagging which will divide the whole review in the words, remove stop words and collect the useful words for negation and conjunction analysis. There major features such as identifying the negation oriented sentiments and the conjunction oriented sentiments which require the analysis of pre-conjunction and post conjunction sentences. So the ambiguity is reduced by analyzing such conjunction and negation based sentences. On analyzed data dual sentiment analysis algorithm is applied which will check the two sides of one review. Finally the polarity of review is checked which will categories the review as positive, negative or neural. By using polarity checking the accuracy of recommendation system is improved. Both the product owner and the user can identify the quality of the product based on the sentiment graph that is generated based on the reviews for each of the product.