Twitter sentiment analysis of the 2019 Indian election

Abstract

With easily available internet services in India and the rest of the world in the recent past, there is more and more traction towards social media like Facebook, Twitter, Instagram, YouTube, etc. This has enabled individuals the freedom of speech and to display their sentiments and emotions towards society. Social media has brought people closer than ever before and has provided a common platform for individuals to communicate. Some influencers promote products on social media platforms, while politicians run their campaigns online for broader reach. Social media has become the fuel for globalization. In 2019, the Indian Lok Sabha Elections saw around 360 million tweets on Twitter, giving their opinions and showing their sentiment towards the political leaders and their parties. Sentiment analysis is the computational investigation of opinions, evaluations, views, and feelings expressed in a text. The political parties have used this technique to run their campaigns and understand the opinions of the public. This also enables them to modify their campaigns accordingly. In this research text mining was performed on approximately 200,000 thousand tweets collected over four months that referenced four national political parties in India during the campaigning period for the Lok Sabha elections in 2019. The sentiments of Twitter users were identified towards each of the considered Indian political parties, Congress, Bhartiya Janata Party (BJP), Aam Aadmi Party (AAP) and Bahujan Samaj Party (BSP) using VADER (Valence Aware Dictionary and sEntiment Reasoner). A lexicon and rule-based sentiment analysis engine was created that is the principal platform to evaluate the opinions expressed in social media. The results of the analysis show that Bhartiya Janata Party (BJP) being a lead runner in the elections of 2019 received more positive intent and emotions towards their campaigns and their leader Narendra Modi as compared to the other parties and their leaders.