Sentiment Analysis of Twitter Data: A Review

Abstract

For a business organization, Sentiment Analysis is a great way to understand the needs of the customers, find out whether they are happy with the product or not and which areas need an improvement. It is also used in various other fields like Healthcare to check how public feels about a specific drug and Politics to predict which party is going to win the election. In light of the current surge in interest in sentiment analysis among researchers for a range of applications, a framework for carrying out a highly accurate sentiment analysis is required. A lot of research has been done for sentiment analysis of tweets from Twitter since it helps in understanding public's view of the particular product, situation or any other topic. There are various models proposed by the researchers with which we can perform sentiment analysis but there are still many challenges that still need to be addressed as well. The presence of emoticons, spam, sarcasm and other forms of content in large number of social media posts makes the feature extraction challenging. Among the Machine Learning Algorithms, the supervised classifiers outperformed the unsupervised approaches in terms of scalability and efficiency. In this review paper, we will discuss the various ways we can perform sentiment analysis and we will also discuss the various applications and challenges of sentiment analysis as well.