From Pixels to Insights: Unsupervised Knowledge Graph Generation with Large Language Model

The role of image data in knowledge extraction and representation has become increasingly significant. This study introduces a novel methodology, termed Image to Graph via Large Language Model (ImgGraph-LLM), which constructs a knowledge graph for each image in a dataset. Unlike existing methods tha...

Full description

Saved in:
Bibliographic Details
Main Authors: Lei Chen, Zhenyu Chen, Wei Yang, Shi Liu, Yong Li
Format: Article
Language:English
Published: MDPI AG 2025-04-01
Series:Information
Subjects:
Online Access:https://www.mdpi.com/2078-2489/16/5/335
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850257782981263360
author Lei Chen
Zhenyu Chen
Wei Yang
Shi Liu
Yong Li
author_facet Lei Chen
Zhenyu Chen
Wei Yang
Shi Liu
Yong Li
author_sort Lei Chen
collection DOAJ
description The role of image data in knowledge extraction and representation has become increasingly significant. This study introduces a novel methodology, termed Image to Graph via Large Language Model (ImgGraph-LLM), which constructs a knowledge graph for each image in a dataset. Unlike existing methods that rely on text descriptions or multimodal data to build a comprehensive knowledge graph, our approach focuses solely on unlabeled individual image data, representing a distinct form of unsupervised knowledge graph construction. To tackle the challenge of generating a knowledge graph from individual images in an unsupervised manner, we first design two self-supervised operations to generate training data from unlabeled images. We then propose an iterative fine-tuning process that uses this self-supervised information, enabling the fine-tuned LLM to recognize the triplets needed to construct the knowledge graph. To improve the accuracy of triplet extraction, we introduce filtering strategies that effectively remove low-confidence training data. Finally, experiments on two large-scale real-world datasets demonstrate the superiority of our proposed model.
format Article
id doaj-art-e9dff79d70a842eebc3c20994c6560d1
institution OA Journals
issn 2078-2489
language English
publishDate 2025-04-01
publisher MDPI AG
record_format Article
series Information
spelling doaj-art-e9dff79d70a842eebc3c20994c6560d12025-08-20T01:56:19ZengMDPI AGInformation2078-24892025-04-0116533510.3390/info16050335From Pixels to Insights: Unsupervised Knowledge Graph Generation with Large Language ModelLei Chen0Zhenyu Chen1Wei Yang2Shi Liu3Yong Li4Department of Electrical Engineering, Tsinghua University, Beijing 100084, ChinaBig Data Center, State Grid Corporation of China, Beijing 100052, ChinaBig Data Center, State Grid Corporation of China, Beijing 100052, ChinaBig Data Center, State Grid Corporation of China, Beijing 100052, ChinaDepartment of Electrical Engineering, Tsinghua University, Beijing 100084, ChinaThe role of image data in knowledge extraction and representation has become increasingly significant. This study introduces a novel methodology, termed Image to Graph via Large Language Model (ImgGraph-LLM), which constructs a knowledge graph for each image in a dataset. Unlike existing methods that rely on text descriptions or multimodal data to build a comprehensive knowledge graph, our approach focuses solely on unlabeled individual image data, representing a distinct form of unsupervised knowledge graph construction. To tackle the challenge of generating a knowledge graph from individual images in an unsupervised manner, we first design two self-supervised operations to generate training data from unlabeled images. We then propose an iterative fine-tuning process that uses this self-supervised information, enabling the fine-tuned LLM to recognize the triplets needed to construct the knowledge graph. To improve the accuracy of triplet extraction, we introduce filtering strategies that effectively remove low-confidence training data. Finally, experiments on two large-scale real-world datasets demonstrate the superiority of our proposed model.https://www.mdpi.com/2078-2489/16/5/335knowledge graphunsupervised learninglarge language model
spellingShingle Lei Chen
Zhenyu Chen
Wei Yang
Shi Liu
Yong Li
From Pixels to Insights: Unsupervised Knowledge Graph Generation with Large Language Model
Information
knowledge graph
unsupervised learning
large language model
title From Pixels to Insights: Unsupervised Knowledge Graph Generation with Large Language Model
title_full From Pixels to Insights: Unsupervised Knowledge Graph Generation with Large Language Model
title_fullStr From Pixels to Insights: Unsupervised Knowledge Graph Generation with Large Language Model
title_full_unstemmed From Pixels to Insights: Unsupervised Knowledge Graph Generation with Large Language Model
title_short From Pixels to Insights: Unsupervised Knowledge Graph Generation with Large Language Model
title_sort from pixels to insights unsupervised knowledge graph generation with large language model
topic knowledge graph
unsupervised learning
large language model
url https://www.mdpi.com/2078-2489/16/5/335
work_keys_str_mv AT leichen frompixelstoinsightsunsupervisedknowledgegraphgenerationwithlargelanguagemodel
AT zhenyuchen frompixelstoinsightsunsupervisedknowledgegraphgenerationwithlargelanguagemodel
AT weiyang frompixelstoinsightsunsupervisedknowledgegraphgenerationwithlargelanguagemodel
AT shiliu frompixelstoinsightsunsupervisedknowledgegraphgenerationwithlargelanguagemodel
AT yongli frompixelstoinsightsunsupervisedknowledgegraphgenerationwithlargelanguagemodel