LPNL: Scalable Link Prediction with Large Language Models

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Bi, Baolong, Liu, Shenghua, Wang, Yiwei, Mei, Lingrui, Cheng, Xueqi
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866910336363266048
author Bi, Baolong
Liu, Shenghua
Wang, Yiwei
Mei, Lingrui
Cheng, Xueqi
author_facet Bi, Baolong
Liu, Shenghua
Wang, Yiwei
Mei, Lingrui
Cheng, Xueqi
contents Exploring the application of large language models (LLMs) to graph learning is a emerging endeavor. However, the vast amount of information inherent in large graphs poses significant challenges to this process. This work focuses on the link prediction task and introduces $\textbf{LPNL}$ (Link Prediction via Natural Language), a framework based on large language models designed for scalable link prediction on large-scale heterogeneous graphs. We design novel prompts for link prediction that articulate graph details in natural language. We propose a two-stage sampling pipeline to extract crucial information from the graphs, and a divide-and-conquer strategy to control the input tokens within predefined limits, addressing the challenge of overwhelming information. We fine-tune a T5 model based on our self-supervised learning designed for link prediction. Extensive experimental results demonstrate that LPNL outperforms multiple advanced baselines in link prediction tasks on large-scale graphs.
format Preprint
id arxiv_https___arxiv_org_abs_2401_13227
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle LPNL: Scalable Link Prediction with Large Language Models
Bi, Baolong
Liu, Shenghua
Wang, Yiwei
Mei, Lingrui
Cheng, Xueqi
Computation and Language
Artificial Intelligence
Machine Learning
Social and Information Networks
Exploring the application of large language models (LLMs) to graph learning is a emerging endeavor. However, the vast amount of information inherent in large graphs poses significant challenges to this process. This work focuses on the link prediction task and introduces $\textbf{LPNL}$ (Link Prediction via Natural Language), a framework based on large language models designed for scalable link prediction on large-scale heterogeneous graphs. We design novel prompts for link prediction that articulate graph details in natural language. We propose a two-stage sampling pipeline to extract crucial information from the graphs, and a divide-and-conquer strategy to control the input tokens within predefined limits, addressing the challenge of overwhelming information. We fine-tune a T5 model based on our self-supervised learning designed for link prediction. Extensive experimental results demonstrate that LPNL outperforms multiple advanced baselines in link prediction tasks on large-scale graphs.
title LPNL: Scalable Link Prediction with Large Language Models
topic Computation and Language
Artificial Intelligence
Machine Learning
Social and Information Networks
url https://arxiv.org/abs/2401.13227