Twitter Sentiment Analysis using Machine Learning and Knowledge-based Approach

Abstract

Sentiment analysis is mainly concerned with identifying and classifying opinions or emotions that are expressed within a text. These days, sharing opinions and expressing emotions through social networking websites has become very common. Therefore, a large amount of data is generated each day, on which mining can be effectively performed to retrieve quality information. Sentiment analysis on such data can prove to be instrumental in generating an aggregated opinion on certain products. Twitter sentiment analysis often becomes a difficult task due to the presence of slangs and misspellings. Also, we constantly encounter new words, which makes it more difficult to analyze and compute the sentiment as compared to the usual sentiment analysis. Twitter restricts the length of a tweet to 140 characters. Thus, extracting valuable information from short texts is yet another challenge. Knowledge-based approach and machine learning can contribute considerably towards the analysis of sentiments from tweets. In this paper, we analyze people's sentiments in their tweets about certain companies. Computing a basic sentiment score and then classifying them as positive or negative would help to serve the company by providing them critical reviews about their products from people worldwide.