Comparative Study of Deep Learning Models for Sentiment Analysis of South Indian Restaurant Reviews
Summary
No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.
Summarise with VIRAT Free to try with your Science Catalogs account.Abstract
Sentiment analysis is an automated methodology for identifying opinions or emotions articulated within textual data.With the proliferation of online platforms for sharing dining experiences, restaurant reviews offer valuable insights for both businesses and consumers.This study concentrates on the sentiment analysis of reviews South Indian restaurants sourced from India, the UAE, and England.Deep learning models including Bidirectional LSTM (BiLSTM) and Bidirectional GRU (BiGRU), both with and without attention mechanisms, are utilized and their performance is compared across GloVe, BERT, and DistilBERT em-beddings.Experimental outcomes indicate that GloVe and BERT embed-dings integrated with BiLSTM or BiGRU models produce high accuracy.The application of attention mechanisms contributes to further perfor-mance gains, with maximum accuracy reaching approximately 95%.The comparative study reinforces the importance of an informed model and embedding choices in domainspecific sentiment analysis.
Fields of Study & Subjects
Sentiment Analysis and Opinion Mining
Topic Modeling
Hate Speech and Cyberbullying Detection
Publication details
- Journal / source
- Advances in intelligent systems research/Advances in Intelligent Systems Research
- Pages
- 939–949
- Publication date
- 1 April 2026
- Language
- EN
- Document type
- Book chapter