KGV-Agent: Schema-Aware Planning and Hybrid Knowledge Toolset for Reliable Knowledge Graph Triple Verification

ACL ARR 2026 January Submission7199 Authors

Keywords: Knowledge Graph， Autonomous Agents ，Triple Verification，Fact-checking，Hybrid Knowledge Fusion， Schema-Aware Planning，Explainable AI，Large Language Models

Abstract: Knowledge Graphs (KGs) serve as a critical foundation for AI systems, yet their automated construction inevitably introduces noise, compromising data trustworthiness. Existing triple verification methods, relying on graph embeddings or pre-trained language models, often suffer from single-source bias—ignoring either internal structural constraints or external semantic context—and adhere to a static inference paradigm. Consequently, they struggle to verify complex or long-tail facts and lack interpretability. To address these limitations, we propose KGV-Agent, a training-free autonomous agent that reframes triple verification as a dynamic process of strategic planning, active investigation, and evidential reasoning. Specifically, KGV-Agent mitigates reasoning instability and cold-start issues via a Memory-Augmented Mechanism and Schema-Aware Strategic Planning. It then executes an enhanced ReAct loop equipped with a Hybrid Knowledge Toolset, dynamically fusing internal structural logic with external textual evidence for robust cross-verification. Empirical results across multiple datasets demonstrate that KGV-Agent establishes new state-of-the-art performance, while providing transparent, fact-based evidence chains for every judgment.

Paper Type: Long

Research Area: NLP Applications

Research Area Keywords: NLP Applications，AI/LLM Agents，Language Modeling，Interpretability and Analysis of Models for NLP，Information Extraction，Information Retrieval and Text Mining，Semantics: Lexical and Sentence-Level

Contribution Types: Model analysis & interpretability, NLP engineering experiment, Reproduction study, Publicly available software and/or pre-trained models

Languages Studied: English

Submission Number: 7199