Detecting sentiment in Twitter data – challenges and implementation

Abstract

Twitter is one of the most popular micro-blogging platforms where users publish their thoughts and opinions and much attention is paid to explore sentiment of these opinions. This paper focuses on the characteristic of Twitter, tweets and supervised machine-learning method for Twitter Sentiment Analysis. Discussion focuses on the following issues: access to the tweets and creating a database, the process of cleaning the database and process of tweets classification into positive and negative groups. The TSA process is presented in Python by simplified architecture.