Selecting personalized gifts is a complex and emotionally significant task, often hindered by uncertainty about recipient preferences—especially in digital contexts lacking direct interaction. This paper introduces GiftMatch, a web-based application that leverages artificial intelligence (AI), user profiling, and conversational interfaces to suggest personalized gifts. Integrating a large language model (LLM)-based API, the system tailors recommendations using data such as interests, age, and personal traits. A systematic literature review explores the evolution of AI-driven recommender systems, with emphasis on generative and conversational methods. Applications like Amazon Gift Finder and Giftster serve as benchmarks. We detail the development of GiftMatch, built with ASP.NET, SQL Server, and the GROQ API. Results show that combining structured data with generative AI enhances user experience in gift selection. The study also discusses challenges in explainability, data privacy, and scalability, contributing insights into the integration of LLMs in emotionally aware, scalable recommendation systems.
Costa et al. (Thu,) studied this question.