From "What to Eat?" to Perfect Recipe: ChefMind's Chain-of-Exploration for Ambiguous User Intent in Recipe Recommendation
This addresses the challenge of fuzzy user demands in recipe recommendation, but it is incremental as it combines existing techniques like Chain of Exploration, Knowledge Graph, RAG, and LLM.
The paper tackled the problem of handling ambiguous user intent in personalized recipe recommendation by proposing ChefMind, a hybrid architecture that achieved an average score of 8.7 versus 6.4-6.7 for baselines and reduced unprocessed queries to 1.6%.
Personalized recipe recommendation faces challenges in handling fuzzy user intent, ensuring semantic accuracy, and providing sufficient detail coverage. We propose ChefMind, a hybrid architecture combining Chain of Exploration (CoE), Knowledge Graph (KG), Retrieval-Augmented Generation (RAG), and a Large Language Model (LLM). CoE refines ambiguous queries into structured conditions, KG offers semantic reasoning and interpretability, RAG supplements contextual culinary details, and LLM integrates outputs into coherent recommendations. We evaluate ChefMind on the Xiachufang dataset and manually annotated queries, comparing it with LLM-only, KG-only, and RAG-only baselines. Results show that ChefMind achieves superior performance in accuracy, relevance, completeness, and clarity, with an average score of 8.7 versus 6.4-6.7 for ablation models. Moreover, it reduces unprocessed queries to 1.6%, demonstrating robustness in handling fuzzy demands.