A review of sentiment analysis techniques for opinionated web text

Social Media nowadays generate huge loads of data that can be valuable in many contexts. It includes media of all formats by which groups of users interact to generate ideas in a distributed and networked process. Data scientists from Twitter have found that the main reason for attaining fame of Presidential Candidate in the upcoming elections scheduled in Nov, 2016 in US is the reach of social media. Researchers and data scientists can use data on social media to track opinions of people about products and services. Many approaches are working behind the scene to reduce errors in opinion mining and sentiment analysis and to attain a level of accuracy for meeting the growing demands of organizations to evaluate their customers. The way people express their opinions have radically changed in the past few years. This paper explores various techniques of distillation of knowledge from huge amount of unstructured information. Generic features of making use of linguistic patterns in sentiment classification are being explored in this paper. In this study investigation of all opinion extraction techniques to generate positive and negative aspects of data with appropriate feature set can help in reduction of error of misclassification.