Data Mining of Social Media for Analysis of Product Review

Social media plays a crucial role in promoting different products. The data collected from the social media helps to improve the quality of products, and helps the customer to select the best product among available products. In this paper, an algorithm is developed based on text mining and TF-IDF (Term Frequency–Inverse Document Frequency) scores. In this paper, it is focused on removing unwanted words such as stop words, stemming words, then the processed data is used for finding sentiment words using NLTK (Natural Language Toolkit). The Stanford POS tagger is also used to tag the words into different categories like positive and negative. The proposed algorithm is implemented using JAVA NetBeans8.2 and achieved desired results. The proposed method can be expanded for the evaluation of different products based on customer reviews provided on the social media.