A Novel Adaptive Approach for Sentiment Analysis on Social Media Data

Abstract

Sentiment analysis (SA) is the approach of determining polarity of any content whether the given sentence contains positive, negative, or neutral sentiments. In many real-world situations, it is required to know the public emotions about happening in surrounding environment. Thus, this analysis helps in decision making on a particular task. There is a huge area where sentiment analysis can be utilized to improve the decision making like while launching a new product, adding additional features in existing products, announcing of a new government policy, etc. This paper shows sentiment analysis system which is based on machine learning algorithm used by TextBlob API using python. Proposed system uses natural language tool kit (NLTK) dataset for training the algorithm. This newly implemented application is used to do sentiment analysis on “twitter” (a social networking application) real-time data. Its experimental results are also presented in this paper. The results/analysis can help big brands, companies, and governments in planning future activities.