< Explain other AI papers

LettinGo: Explore User Profile Generation for Recommendation System

Lu Wang, Di Zhang, Fangkai Yang, Pu Zhao, Jianfeng Liu, Yuefeng Zhan, Hao Sun, Qingwei Lin, Weiwei Deng, Dongmei Zhang, Feng Sun, Qi Zhang

2025-06-24

LettinGo: Explore User Profile Generation for Recommendation System

Summary

This paper talks about LettinGo, a system that creates diverse and flexible user profiles using large language models to make recommendations better suited to individual preferences.

What's the problem?

The problem is that traditional recommendation systems often rely on limited or static user information, which can lead to less accurate and less personalized suggestions.

What's the solution?

The researchers developed LettinGo, which generates multiple user profiles that can adapt over time by using advanced language models and a method called Direct Preference Optimization to better capture user interests and changes.

Why it matters?

This matters because it helps recommendation systems understand users more deeply and provide more accurate, personalized suggestions, improving user satisfaction in apps, websites, and online services.

Abstract

LettinGo enhances user profiling via diverse, adaptive profiles generated using LLMs and Direct Preference Optimization, improving recommendation accuracy and flexibility.