Sentiment Classification of Safaricom PLC Social Media Sentiments on X(Formerly Twitter)
Joshua Kimani *
Department of Mathematics, Multimedia University of Kenya, Kenya.
Anthony Karanjah
Department of Mathematics, Multimedia University of Kenya, Kenya.
Pius Kihara
Department of Financial and Actuarial Mathematics, Technical University of Kenya, Kenya.
*Author to whom correspondence should be addressed.
Abstract
In today's digital era, social media plays a pivotal role in shaping public sentiment, particularly in the financial domain. This study focuses on sentiment analysis of social media discussions, specifically tweets discussing Safaricom PLC on X (formerly Twitter), leveraging Natural Language Processing (NLP) techniques. By meticulously collecting, cleaning, and analyzing data, valuable insights into the sentiment landscape surrounding Safaricom PLC during a significant period were obtained. The sentiment analysis, conducted using the VADER lexicon, categorized sentiments into positive, negative, and neutral classes. Notably, the analysis revealed a predominant positive sentiment, indicative of an optimistic tone in discussions related to Safaricom PLC. This study highlights the potential of integrating sentiments and sentiment analysis techniques into stock price prediction models to facilitate informed investment decision-making.
Keywords: Sentiment analysis, social media, Safaricom PLC, stock price prediction, language processing (NLP), VADER lexicon