Diagnostic-Guided Dynamic Profile Optimization for LLM-based User Simulators in Sequential Recommendation

TL;DR

This paper introduces DGDPO, a dynamic, iterative framework for constructing realistic user profiles in LLM-based simulators, improving accuracy and multi-round interaction fidelity in sequential recommendation scenarios.

Contribution

It presents a novel diagnostic-guided, iterative profile optimization method that enhances LLM-based user simulators for more realistic, multi-round recommendation interactions.

Findings

Improved user profile accuracy in simulations.

Enhanced multi-round recommendation interaction fidelity.

Demonstrated effectiveness on real-world datasets.

Abstract

Recent advances in large language models (LLMs) have enabled realistic user simulators for developing and evaluating recommender systems (RSs). However, existing LLM-based simulators for RSs face two major limitations: (1) static and single-step prompt-based inference that leads to inaccurate and incomplete user profile construction; (2) unrealistic and single-round recommendation-feedback interaction pattern that fails to capture real-world scenarios. To address these limitations, we propose DGDPO (Diagnostic-Guided Dynamic Profile Optimization), a novel framework that constructs user profile through a dynamic and iterative optimization process to enhance the simulation fidelity. Specifically, DGDPO incorporates two core modules within each optimization loop: firstly, a specialized LLM-based diagnostic module, calibrated through our novel training strategy, accurately identifies specific defects in the user profile. Subsequently, a generalized LLM-based treatment module analyzes the diagnosed defect and generates targeted suggestions to refine the profile. Furthermore, unlike existing LLM-based user simulators that are limited to single-round interactions, we are the first to integrate DGDPO with sequential recommenders, enabling a bidirectional evolution where user profiles and recommendation strategies adapt to each other over multi-round interactions. Extensive experiments conducted on three real-world datasets demonstrate the effectiveness of our proposed framework.