Sentiment Analysis on Yelp Reviews with BERT and RoBERTa

Project Overview

This repository contains the code and documentation for performing sentiment analysis on Yelp reviews using two state-of-the-art models: Hugging Face's bert-base-multilingual-uncased and cardiffnlp/twitter-roberta-base-sentiment-latest. The project aims to compare these models' effectiveness in analyzing sentiments expressed in Yelp reviews.

Access our models through this link

https://drive.google.com/drive/folders/1CyRpe-uADqTpcOlKk7TLk4POsj3nmt4u?usp=sharing

Models Used

• BERT Multilingual Uncased: A model capable of understanding multiple languages, ideal for the diverse linguistic nature of Yelp reviews.
• Twitter RoBERTa: Specially fine-tuned for sentiment analysis, this model excels in understanding the nuances of English-language sentiments.

Getting Started

Prerequisites

requirements.txt

Installation

1. Clone the repository:

   git clone https://github.com/oshengeenath/Sentiment_Analysis_on_Yelp_Reviews_Dataset_with_BERT_and_RoBERTa

2. Install required Python packages:

   pip install -r requirements.txt

Acknowledgements

• Hugging Face for the transformer models. Link to the model -> https://huggingface.co/cardiffnlp/twitter-roberta-base-sentiment-latest
• CardiffNLP for the Twitter RoBERTa model. Link to the model -> https://huggingface.co/nlptown/bert-base-multilingual-uncased-sentiment
• Yelp for the dataset. Link to the Yelp dataset -> https://www.yelp.com/dataset