• Kunal Shrivastava
• Aditya Pathak
• Shashank Shekhar Singh
• Sarveshvar Mishra
• Gagan Kumar Singh

This study presents a hybrid sentiment analysis model that combines VADER (a fast, rule-based tool for short texts) with Hugging Face’s transformer-based models (for contextual accuracy). Using a dataset of 150,000 product reviews, the hybrid approach assigns VADER to short reviews and BERT to longer or ambiguous ones. The model achieves 95% accuracy with low latency (0.06s), outperforming baseline, VADER-only, and BERT-only models. This efficient and scalable framework is ideal for real-time e-commerce applications like customer feedback analysis.

Core Technologies & Tools: VADER Hugging Face BERT Transformers NLP (Natural Language Processing) Machine Learning Deep Learning Sentiment Analysis Lexicon-based model Transformer-based model Hybrid Model Logistic Regression Ensemble Method Fine-tuning Tokenization Preprocessing Application Domain: Product Reviews E-commerce Customer Feedback Real-time Sentiment Analysis Opinion Mining Review Classification Text Analytics Metrics & Evaluation: Accuracy Precision Recall F1-Score AUC (Area Under Curve) Latency Robustness Dataset & Text Characteristics: Crowdsourced Labeling Review Length Short Texts Long Texts Informal Language Emojis Slang Noise Injection Models & Algorithms: Logistic Regression Naive Bayes Support Vector Machines (SVM) LSTM (Long Short-Term Memory) mBERT (Multilingual BERT) Enhancement Techniques: Weighted Ensemble Dynamic Weighting Aspect-Based Sentiment Analysis Edge Device Deployment Multilingual Support

"Enhancing Sentiment Analysis of Product Reviews Using VADER and Hugging Face", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.12, Issue 5, page no.f570-f575, May-2025, Available :http://www.jetir.org/papers/JETIR2505665.pdf

"Enhancing Sentiment Analysis of Product Reviews Using VADER and Hugging Face", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.12, Issue 5, page no. ppf570-f575, May-2025, Available at : http://www.jetir.org/papers/JETIR2505665.pdf