Shengguang Bai

Dynamic path generation in complex transportation networks is essential for intelligent transportation systems. Traditional methods, such as shortest path algorithms or heuristic-based models, often fail to capture real-world travel behaviors due to their reliance on simplified assumptions and limited ability to handle long-range dependencies or non-linear patterns. To address these limitations, we propose PathGen-LLM, a large language model (LLM) designed to learn spatial–temporal patterns from historical paths without requiring handcrafted features or graph-specific architectures. Exploiting the structural similarity between path sequences and natural language, PathGen-LLM converts spatiotemporal trajectories into text-formatted token sequences by encoding node IDs and timestamps. This enables the model to learn global dependencies and semantic relationships through self-supervised pretraining. The model integrates a hierarchical Transformer architecture with dynamic constraint decoding, which synchronizes spatial node transitions with temporal timestamps to ensure physically valid paths in large-scale road networks. Experimental results on real-world urban datasets demonstrate that PathGen-LLM outperforms baseline methods, particularly in long-distance path generation. By bridging sequence modeling and complex network analysis, PathGen-LLM offers a novel framework for intelligent transportation systems, highlighting the potential of LLMs to address challenges in large-scale, real-time network tasks.

Subway origin–destination (OD) demand generation involves estimating passenger travel demand between stations by leveraging geographic data and the structure of the subway network. This process is essential for effective urban subway system planning and development, substantially contributing to enhanced system efficiency and reduced urban congestion. Recent developments in subway feeder modes, notably bike-sharing systems, have significantly altered subway passengers’ access and egress travel behaviors, expanded station service coverage and introduced additional complexity into OD demand generation tasks. However, the existing methodologies often overlook station service areas (SSAs) and fail to provide a comprehensive macro-level perspective of the entire subway network. To address these gaps, this paper proposes a two-stage graph-based OD demand generation model that integrates an SSA prediction module and a graph attention network (GAT). This integrated framework facilitates the accurate estimation of inter-station travel demands over varying time intervals. Experimental evaluations conducted using Beijing subway data demonstrate that the proposed model consistently outperforms established baseline approaches, highlighting the critical roles of both the SSA prediction module and the GAT in achieving improved performance.

Tianyang Zhang, Zhuoxuan Jiang, Shengguang Bai, Tianrui Zhang, Lin Lin, Yang Liu, Jiawei Ren. Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing: Industry Track. 2024.