Research

Paper

TESTING February 27, 2026

MMKG-RDS: Reasoning Data Synthesis via Deep Mining of Multimodal Knowledge Graphs

Authors

Lun Zhan, Feng Xiong, Huanyong Liu, Feng Zhang, Yuhui Yin

Abstract

Synthesizing high-quality training data is crucial for enhancing domain models' reasoning abilities. Existing methods face limitations in long-tail knowledge coverage, effectiveness verification, and interpretability. Knowledge-graph-based approaches still fall short in functionality, granularity, customizability, and evaluation. To address these issues, we propose MMKG-RDS, a flexible framework for reasoning data synthesis that leverages multimodal knowledge graphs. It supports fine-grained knowledge extraction, customizable path sampling, and multidimensional data quality scoring. We validate MMKG-RDS with the MMKG-RDS-Bench dataset, covering five domains, 17 task types, and 14,950 samples. Experimental results show fine-tuning Qwen3 models (0.6B/8B/32B) on a small number of synthesized samples improves reasoning accuracy by 9.2%. The framework also generates distinct data, challenging existing models on tasks involving tables and formulas, useful for complex benchmark construction. The dataset and code are available at https://github.com/360AILAB-NLP/MMKG-RDS

Metadata

arXiv ID: 2602.23632
Provider: ARXIV
Primary Category: cs.AI
Published: 2026-02-27
Fetched: 2026-03-02 06:04

Related papers

Raw Data (Debug)
{
  "raw_xml": "<entry>\n    <id>http://arxiv.org/abs/2602.23632v1</id>\n    <title>MMKG-RDS: Reasoning Data Synthesis via Deep Mining of Multimodal Knowledge Graphs</title>\n    <updated>2026-02-27T03:10:51Z</updated>\n    <link href='https://arxiv.org/abs/2602.23632v1' rel='alternate' type='text/html'/>\n    <link href='https://arxiv.org/pdf/2602.23632v1' rel='related' title='pdf' type='application/pdf'/>\n    <summary>Synthesizing high-quality training data is crucial for enhancing domain models' reasoning abilities. Existing methods face limitations in long-tail knowledge coverage, effectiveness verification, and interpretability. Knowledge-graph-based approaches still fall short in functionality, granularity, customizability, and evaluation. To address these issues, we propose MMKG-RDS, a flexible framework for reasoning data synthesis that leverages multimodal knowledge graphs. It supports fine-grained knowledge extraction, customizable path sampling, and multidimensional data quality scoring. We validate MMKG-RDS with the MMKG-RDS-Bench dataset, covering five domains, 17 task types, and 14,950 samples. Experimental results show fine-tuning Qwen3 models (0.6B/8B/32B) on a small number of synthesized samples improves reasoning accuracy by 9.2%. The framework also generates distinct data, challenging existing models on tasks involving tables and formulas, useful for complex benchmark construction. The dataset and code are available at https://github.com/360AILAB-NLP/MMKG-RDS</summary>\n    <category scheme='http://arxiv.org/schemas/atom' term='cs.AI'/>\n    <published>2026-02-27T03:10:51Z</published>\n    <arxiv:primary_category term='cs.AI'/>\n    <author>\n      <name>Lun Zhan</name>\n    </author>\n    <author>\n      <name>Feng Xiong</name>\n    </author>\n    <author>\n      <name>Huanyong Liu</name>\n    </author>\n    <author>\n      <name>Feng Zhang</name>\n    </author>\n    <author>\n      <name>Yuhui Yin</name>\n    </author>\n  </entry>"
}