Project Overview
- Built NLP pipelines for spam detection, sentiment classification, and complaint resolution tasks
- Utilized word embeddings (GloVe, FastText) to capture semantic meaning and reduce dimensionality
- Trained LSTM-based sequence models followed by feedforward layers for binary classification
- Performed text preprocessing including tokenization, padding, and sequence modeling
Repositories
📋 Consumer Complaints Resolution
NLP pipeline to classify and route consumer complaints to appropriate departments using text classification techniques.
View on GitHub →⭐ Amazon Cell Labels Sentiment Prediction
Sentiment analysis model for Amazon product reviews using deep learning to classify positive/negative sentiments.
View on GitHub →🔍 Quora Spam Filter Prediction
Spam detection system using GloVe and FastText word embeddings with LSTM networks to identify insincere questions.
View on GitHub →