Schema-Aware Planning and Hybrid Knowledge Toolset for Reliable Knowledge Graph Triple Verification
Summary: arXiv:2604.04190v1 Announce Type: new
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, based on graph embeddings or language models, often suffer from single-source bias by relying on either internal structural constraints or external semantic evidence, and usually follow a static inference paradigm. As a result, they struggle with complex or long-tail facts and provide limited interpretability.
To address these limitations, we propose SHARP (Schema-Hybrid Agent for Reliable Prediction), a training-free autonomous agent that reformulates triple verification as a dynamic process of strategic planning, active investigation, and evidential reasoning.
Key Features of SHARP
- Memory-Augmented Mechanism: This feature enhances reasoning stability by allowing the agent to utilize past information effectively.
- Schema-Aware Strategic Planning: This approach facilitates a more informed decision-making process that considers both the internal and external environments of the knowledge graph.
- Enhanced ReAct Loop: The ReAct loop is optimized to dynamically integrate internal KG structure and external textual evidence for accurate cross-verification.
- Hybrid Knowledge Toolset: This toolset allows SHARP to leverage multiple sources of information, improving the reliability of its predictions.
Experimental Results
Experiments conducted on FB15K-237 and Wikidata5M-Ind datasets indicate that SHARP significantly outperforms existing state-of-the-art baselines. The results show accuracy gains of 4.2% and 12.9%, respectively, highlighting SHARP’s effectiveness in handling complex verification tasks.
Interpretability and Robustness
One of the standout features of SHARP is its capacity to provide transparent, fact-based evidence chains for each judgment made. This not only enhances the interpretability of the verification process but also demonstrates robustness when dealing with complex verification tasks. By offering clear evidence chains, SHARP allows users to trace back the reasoning process, making it easier to understand the basis for each outcome.
Conclusion
In summary, SHARP represents a significant advancement in the field of knowledge graph triple verification. By combining strategic planning with a hybrid approach to knowledge integration, SHARP addresses many of the shortcomings of traditional methods. Its dynamic, autonomous framework sets a new standard for reliability and interpretability in knowledge graph applications, paving the way for more trustworthy AI systems.