Amazon Reviews Sentiment Analysis

Used dataset

Project relevance:

To explore existing sentiment analysis approaches, understand how they differ, and evaluate their impact on sentiment analysis performance for Amazon reviews

Research aims:

1. Gain insights into the nature of unlabeled reviews data
2. Explore the potential efficiency of clustering algorithms on text data
3. Investigate strategies for accurate sentiment analysis implementation for chosen data

Tools and techniques:

pandas, numpy, matplotlib, seaborn, nltk, scikit-learn, torch, RoBERTa transformer model