Published Paper Details:

The rapid expansion of online shopping platforms has significantly increased the variety of products available to users, making the decision-making process more complex and time-consuming. Conventional recommendation systems, such as collaborative filtering and content-based approaches, often struggle to accurately interpret user intent expressed in natural language and typically lack transparency in their outputs. This paper presents a novel design of an AI-driven explainable recommendation system that integrates Large Language Models (LLMs), semantic similarity search using FAISS, and interpretable machine learning techniques such as SHAP. The proposed system is capable of processing user queries in natural language, extracting meaningful preferences, and retrieving contextually relevant products from large datasets. Additionally, it generates clear and human-understandable explanations for each recommendation. By combining advanced language understanding with explainability and efficient retrieval mechanisms, the system improves recommendation accuracy, enhances transparency, and increases user trust. The proposed framework demonstrates the potential of integrating modern AI techniques to build intelligent and usercentric recommendation systems. Index Terms--Explainable Artificial Intelligence (XAI), Recommendation Systems, Large Language Models (LLMs), Semantic Search, FAISS, SHAP, Natural Language Processing (NLP), E-commerce, Personalized Recommendations, Information Retrieval