SENTIMENT ANALYSIS USING TRANSFORMER-BASED MODELS

Authors

  • Dr. Ayesha Farooq Department of Computer Science, University of Karachi, Pakistan.

DOI:

https://doi.org/10.71465/csb3

Keywords:

Sentiment analysis, transformer models, BERT, NLP, contextual embedding’s, IMDB, Twitter Sentiment140

Abstract

Sentiment analysis has become an essential task in natural language processing (NLP), facilitating applications in marketing, social media monitoring, and customer service. Transformer-based models, such as BERT, RoBERTa, and GPT, have significantly advanced the state-of-the-art in this domain by leveraging deep contextual embedding’s and self-attention mechanisms. This paper explores the application of transformer-based architectures to sentiment analysis tasks, highlighting their advantages over traditional methods. By comparing model performance on datasets such as IMDB and Twitter Sentiment140, this study demonstrates the superior accuracy and efficiency of transformer models. The paper also discusses challenges, including computational demands and data pre-processing, offering strategies to optimize deployment.

image

Downloads

Published

2018-06-30