Implementation of an interactive chatbot for summarizing legal and policy documents. Includes data preprocessing (cleaning, tokenization, hierarchical chunking), extractive TF-IDF baselines, and fine-tuned abstractive models (DistilBART, LED). Integrates a retrieval layer for document relevance and uses ROUGE, BLEU, and cosine similarity metrics.

Exploration of retrieval methods on the HotpotQA corpus, combining dense retrieval and feature-based reranking. Achieved a mean nDCG@10 of 0.9416 using LambdaRank with features such as cross-encoder score, LLM score, BM25 score, and token-based statistics—surpassing dense retriever + cross-encoder baselines.

This repository contains implementations and workflows for key NLP tasks like text classification, Generative AI, sentiment analysis, and entity recognition. It includes preprocessing scripts, annotated datasets, and fine-tuning methods for frameworks like Hugging Face and spaCy. Ideal for building and deploying scalable NLP solutions.