Design and implementation of a framework to aggregate feedbacks about consumer products from social media

With the growth in business’s competitive environment, organizations are obliged in tracking consumer’s data so that they can gain success in their business’s aspects. With the advancement in technology of 21st century, a new phase of business competition has started.  This phase is to use online data to get more information about the market trends and the customer’s reviews, so that organizations might gain edge over each other. The aim of this research is to help achieve that goal by analyzing the user’s opinion about the specific product on social media, classifying if it is gaining positive or negative remarks, providing the consumer with said information and also providing them information about related products that they might want. We collect a dataset from three different social media platforms i.e., Facebook, Twitter and Amazon. This dataset includes reviews and user comments about products and their ratings. We took six attributes into account for research purpose. The research techniques we applied on these three datasets are Naive Bayesian algorithm and Vader sentiment  algorithm  for sentiment analysis. We then applied chi-squared algorithm for weight calculation. The weight is then applied to attributes and resultant calculations were made. We analyze the results both in mathematical and graphical form, allowing the customers to review overall feedback of consumers on products. We then devised a framework which uses both offline and online repositories. This research helps us in understanding the user’s behavior towards individual products and what can be done to improve them or which product is best for the market use.