How Can Large Language Models Understand Spatial-Temporal Data?

TL;DR

This paper introduces STG-LLM, a novel method that adapts large language models for spatial-temporal forecasting by transforming graph data into tokens and fine-tuning minimal parameters, achieving state-of-the-art results.

Contribution

The paper proposes STG-Tokenizer and STG-Adapter to enable LLMs to understand and forecast spatial-temporal data effectively, bridging the gap between graph data and language models.

Findings

Achieves competitive performance on spatial-temporal benchmarks.

Successfully adapts LLMs for complex graph data.

Maintains natural language understanding while forecasting.

Abstract

While Large Language Models (LLMs) dominate tasks like natural language processing and computer vision, harnessing their power for spatial-temporal forecasting remains challenging. The disparity between sequential text and complex spatial-temporal data hinders this application. To address this issue, this paper introduces STG-LLM, an innovative approach empowering LLMs for spatial-temporal forecasting. We tackle the data mismatch by proposing: 1) STG-Tokenizer: This spatial-temporal graph tokenizer transforms intricate graph data into concise tokens capturing both spatial and temporal relationships; 2) STG-Adapter: This minimalistic adapter, consisting of linear encoding and decoding layers, bridges the gap between tokenized data and LLM comprehension. By fine-tuning only a small set of parameters, it can effectively grasp the semantics of tokens generated by STG-Tokenizer, while preserving the original natural language understanding capabilities of LLMs. Extensive experiments on diverse spatial-temporal benchmark datasets show that STG-LLM successfully unlocks LLM potential for spatial-temporal forecasting. Remarkably, our approach achieves competitive performance on par with dedicated SOTA methods.