0
- Volume 2, Issue 2 2024
By Kulsoom Nasir Hussain* Maria Andleeb†
Keywords: sentiment analysis, machine learning, social media, twitter, reddit
Sentiment analysis has been a topic of intense research. This study performed sentiment analysis to explore user perception and attitude towards ChatGPT using sentiment analysis. In this study, sentiment analysis techniques are applied to two datasets. These datasets include i.e. tweets and Reddit posts. It contains references to ChatGPT within the context of education. The study aims to reveal the user’s sentiment, opinions, and comments expressed on these social media platforms regarding the use of ChatGPT in educational environments. The study adopts the following methodology. A diverse quantity of textual data is collected in real time from Twitter and Reddit. The collected data is preprocessed to improve its quality and to extract valuable information. A whole natural language processing pipeline is applied such as tokenization, stemming, lemmatization, the removal of stop words, normalization, and replacement of similar words with a common term. The textual data is classified into a sentiment category by two different sentiment analysis models i.e. Logistic regression and support vector machines. In addition, this study performed the sentiment analysis using lexicon-based approaches employing different NLP tools.