TrajAgent: An LLM-Agent Framework for Trajectory Modeling via Large-and-Small Model Collaboration

TL;DR

TrajAgent presents an LLM-driven agent framework for automated, unified trajectory modeling across heterogeneous data and tasks. It couples UniEnv, an execution environment for diverse trajectory data and models, with an agentic workflow and a collaborative learning schema that bridges high-level reasoning and low-level model training. The approach yields consistent performance gains across five core trajectory tasks and four real-world datasets, and mitigates optimization failures through memory pruning and contrastive reflection. This framework enables robust, automated trajectory modeling with transferable improvements across tasks and datasets, and its codebase is publicly accessible. The work has practical implications for scalable, cross-domain trajectory analytics in urban planning, mobility services, and location-based applications.

Abstract

Trajectory modeling, which includes research on trajectory data pattern mining and future prediction, has widespread applications in areas such as life services, urban transportation, and public administration. Numerous methods have been proposed to address specific problems within trajectory modeling. However, the heterogeneity of data and the diversity of trajectory tasks make effective and reliable trajectory modeling an important yet highly challenging endeavor, even for domain experts. In this paper, we propose TrajAgent, an agent framework powered by large language models, designed to facilitate robust and efficient trajectory modeling through automation modeling. This framework leverages and optimizes diverse specialized models to address various trajectory modeling tasks across different datasets effectively. In TrajAgent, we first develop UniEnv, an execution environment with a unified data and model interface, to support the execution and training of various models. Building on UniEnv, we introduce an agentic workflow designed for automatic trajectory modeling across various trajectory tasks and data. Furthermore, we introduce collaborative learning schema between LLM-based agents and small speciallized models, to enhance the performance of the whole framework effectively. Extensive experiments on five tasks using four real-world datasets demonstrate the effectiveness of TrajAgent in automated trajectory modeling, achieving a performance improvement of 2.38%-69.91% over baseline methods. The codes and data can be accessed via https://github.com/tsinghua-fib-lab/TrajAgent.

Figures (5)

• Figure 1: The paradigm of LLM based automated trajectory modeling framework TrajAgent.
• Figure 2: The whole framework of TrajAgent.
• Figure 3: The TrajAgent framework supports 5 fundamental trajectory modeling tasks, encompassing a total of 18 methods. Detailed introduction of methods can refer to the appendix \ref{['sec:appendix']}.
• Figure 4: (a-b) The impact of thought steps and memory size when using DeepMove in next location prediction tasks. (c) Compared to Optuna and UrbanLLM, TrajAgent achieves better performance in trajectory user linking tasks with S2TUL as the specialized model.
• Figure 5: A representative example case of TrajAgent.