End-to-End Graph Flattening Method for Large Language Models

TL;DR

This paper introduces EEDP, a novel graph flattening method for large language models that improves reasoning over long-distance graph data while maintaining performance in short-distance scenarios.

Contribution

The paper proposes an end-to-end graph flattening technique inspired by human reasoning, enhancing LLMs' ability to understand long-distance graph relationships.

Findings

EEDP improves reasoning in long-distance graph scenarios.

EEDP maintains high performance in short-distance scenarios.

EEDP demonstrates robustness to distance variations.

Abstract

In recent years, the breakthrough of Large Language Models (LLMs) offers new ideas for achieving universal methods on graph data. The common practice of converting graphs into natural language for LLMs, which refers to graph flattening, exhibits good generalizability and interpretability. However, the poor organization of the textual format results in poor performance in long-distance scenario understanding. Inspired by human cognitive reasoning habits, we propose a novel method for graph flattening to fit LLMs, termed as End-to-End DAG-Path prompting (EEDP). Experiments on real-world datasets show that EEDP enhances the reasoning performance of LLMs in long-distance scenarios while maintaining excellent performance in short-distance scenarios, demonstrating good robustness in the face of distance variations.