Abstract:
The retail clothing sector continues to face significant challenges in delivering seamless and personalized customer service, particularly during peak shopping periods when human support is limited and response times increase. Existing chatbot solutions are largely based on rule-driven logic or traditional machine learning and Natural Language Understanding (NLU) techniques, which restrict their ability to interpret complex customer queries and generate dynamic, context-aware responses. As a result, these systems often fail to provide accurate fashion guidance, personalized product recommendations, and smooth end-to-end support from product discovery to purchase, ultimately reducing customer engagement and potential sales.
This research proposes and implements an AI-powered shopping assistant tailored specifically for clothing retail, using a hybrid intelligent architecture. The system combines NLU for intent detection and basic query handling, Retrieval-Augmented Generation (RAG) to accurately answer store policies, FAQs, and product-related questions using a curated knowledge base, and a Large Language Model (LLM) enhanced through domain-specific prompt engineering to manage complex, conversational, and advisory interactions. In addition, a recommendation engine is integrated using both content-based and collaborative filtering techniques to generate personalized fashion suggestions based on user preferences, behavior, and shopping history.
The assistant is designed to support natural language conversations, real-time fashion advice, and guided shopping journeys, enabling users to interact using simple everyday language. System performance was evaluated using response accuracy, relevance, user satisfaction, and engagement metrics. Experimental results show that the proposed hybrid assistant outperforms traditional chatbot systems by delivering more accurate, context-aware, and personalized responses. The study demonstrates that combining RAG, LLMs, and intelligent recommendation methods can significantly enhance digital retail experiences and provides a scalable, user-friendly solution for next-generation clothing store assistants.
