RecoMob: Opinion mining for product enhancement

Opinion mining is the field of extracting and studying people's' opinions, sentiments, attitudes, and emotions expressed in a different digital form in the form of reviews on e-commerce and other social networking sites. It has a broad range of applications which varies from understanding the personal likes and dislikes of the particular user to predicting his shopping habits. The prime objective of our project is to understand the some of the features of the mobile based on the reviews given by the users, classify them into positive and negative, and also suggest the products to the user based on the parameters specified by the user. To implement this, we have obtained the reviews by scrapping the Amazon website for six phone models and later cleaning the data contains sentiments of people which are expressed in different ways. Our application will collect reviews about products, calculate the sentiment score of the words and hence determine the overall polarity of the sentence as positive or negative. We have developed a dashboard wherein the customer can easily identify the specific features of the mobile. We have also predicted the similar products based on the user requirements. The graphs which indicate the popularity of the product over time. The final output of the system will be - 1) Timeline Analysis of product reviews 2) Product Recommendation.