Sentiment Analysis Using Twitter
Analysing the sentiment from the tweets of a user can give general insights into how the person is thinking. This paper analyses different groups of people like graduate students, politicians and doctors and aims to study the behavioural patterns. A metho
- PDF / 295,691 Bytes
- 8 Pages / 439.37 x 666.142 pts Page_size
- 23 Downloads / 227 Views
Abstract Analysing the sentiment from the tweets of a user can give general insights into how the person is thinking. This paper analyses different groups of people like graduate students, politicians and doctors and aims to study the behavioural patterns. A method was designed to extract sentiment from tweets obtained using Twitter API. The VADER sentiment analysis library was used to gather median sentiment scores for different groups of people. The results show that politicians have a similar propensity for positive and negative tweets. Comedians’ tweets have the best positive median score, while graduate students’ tweets have the least positive median score among the tested groups. This paper extensively compares tweets of different groups of people and provides a novel analysis of general sentiment trends. Keywords Sentiment analysis · VADER · Twitter
1 Introduction Sentiment analysis is a machine learning approach to gather knowledge about the emotional state of an author from their authored piece of text. This paper utilizes the VADER (Valence Aware Dictionary and sentiment Reasoner) sentiment analysis library to analyse the sentiment of different groups of people, namely politicians, doctors, comedians, motivational speakers and graduate students. The VADER library was specially developed for social media analysis. It is a simple rule-based model that is used to analyse typical social content using qualitative and quantitative methods. Sentiment analysis is used in various fields like marketing research, customer psychology studies, economics and business strategy development. U. Bhaumik (B) · D. K. Yadav Katihar Engineering College, University College London (UK), Katihar, India e-mail: [email protected] D. K. Yadav e-mail: [email protected] © The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2021 J. K. Mandal et al. (eds.), Computational Intelligence and Machine Learning, Advances in Intelligent Systems and Computing 1276, https://doi.org/10.1007/978-981-15-8610-1_7
59
60
U. Bhaumik and D. K. Yadav
Twitter is a popular social media that focuses mainly on current world affairs and also allows users to be connected by allowing them to follow each other and also retweet other’s tweets [1]. It is a valuable source of data as millions of users use the platform every day. The tweets are generally publicly available can be downloaded using Twitter’s application programming interface. As a result, this makes it possible to study people’s opinions, holistically and on the individual level too [1]. The social Twitter Chirp developer conference in April 2010 [2] provided some statistics about Twitter and the platform’s user engagement. According to Twitter, 3 lakh new users were added to the site every day and over 600 million queries populated its search engine, and above 3 billion requests per day were carried out based on the results of Twitter API. Nearly 40% of the users used their mobile to tweet. As of September 2019, Twit
Data Loading...
