---
title: "HLMTrans: A Sim-to-Real Transfer Framework for Spatial Crowdsourcing with Human-Guided Language Models"
authors:
  - "Qingshun Wu"
  - "Yafei Li"
  - "Lulu Li"
  - "Yuanyuan Jin"
  - "Shuo He"
  - "Mingliang Xu"
doi: "10.24963/ijcai.2025/471"
url: "https://ijcai.org/proceedings/2025/471"
pdf: "https://ijcai.org/proceedings/2025/0471.pdf"
venue: "Proceedings of the Thirty-Fourth International Joint Conference on Artificial Intelligence"
year: 2025
pages: "4227-4235"
track: "Main Track"
keywords:
  - "Humans and AI: HAI: Applications"
  - "Humans and AI: HAI: Human computation and crowdsourcing"
  - "Planning and Scheduling: PS: Real-time planning"
  - "Machine Learning: ML: Applications"
---

# HLMTrans: A Sim-to-Real Transfer Framework for Spatial Crowdsourcing with Human-Guided Language Models

Qingshun Wu, Yafei Li, Lulu Li, Yuanyuan Jin, Shuo He, Mingliang Xu

*Proceedings of the Thirty-Fourth International Joint Conference on Artificial Intelligence*, pages 4227-4235. DOI: [10.24963/ijcai.2025/471](https://doi.org/10.24963/ijcai.2025/471)

## Abstract

Reinforcement Learning (RL), trained via trial and error in simulators, has been proven to be an effective approach for addressing task assignment problems in spatial crowdsourcing. However, a performance gap still exists when transferring the simulator-trained RL Models (RLMs) to real-world settings due to the misalignment of travel time. Existing works mostly focus on using data-driven and learning-based methods to predict travel time; unfortunately, these approaches are limited in achieving accurate predictions by requiring a large amount of real-world data covering the entire state distribution. In this paper, we propose a Sim-to-Real Transfer with Human-guided Language Models framework called HLMTrans, which comprises three core modules: RLMs decision for task assignment, sim-to-real transfer with Large Language Models (LLMs), and preference learning from human feedback. HLMTrans first leverages the zero-shot chain-of-thought reasoning capability of LLMs to estimate travel time by capturing the real-world dynamics. This estimation is then input as domain knowledge into the forward model of Grounded Action Transformation (GAT) to enhance the action transformation of RLMs. Further, we design a human preference learning mechanism to fine-tune LLMs, improving their generation quality and enabling RLMs learn a more realistic policy. We evaluate the proposed HLMTrans on two real-world datasets, and the experimental results demonstrate that HLMTrans outperforms the SOTA methods in terms of effectiveness and efficiency.