Improved Tweets Polarity Detection using Lexicon-based Features and Caching

Abstract

Tweet Polarity detection the process of observing and identifying the sentiment inclination of text, whether it is positive or negative. In this paper, improved polarity detection on tweets using supervised learning is proposed. This method is using data sets available in public. The pre-processing is improved using proper caching of data items to save the time for processing of duplicate items in data sets. The feature selection strategy ensures reduced dimensionality. The low dimension data improves the classification efficiency. The experiment shows that the method is improving the overall performance in training and testing of polarity detection.