Adjudicator: Correcting Noisy Labels with a KG-Informed Council of LLM Agents
This addresses the critical issue of data quality for high-stakes industrial applications, though it appears incremental as it builds on existing neuro-symbolic and multi-agent approaches.
The paper tackles the problem of noisy labels in production machine learning systems by introducing Adjudicator, a system that uses a knowledge graph and a multi-agent LLM council to correct labels, achieving a 0.99 F1-score on a benchmark dataset.
The performance of production machine learning systems is fundamentally limited by the quality of their training data. In high-stakes industrial applications, noisy labels can degrade performance and erode user trust. This paper presents Adjudicator, a system that addresses the critical data mining challenge of automatically identifying and correcting label noise and has been validated for production deployment. Adjudicator models this as a neuro-symbolic task, first constructing a dynamic Knowledge Graph (KG) to unify item context. This KG then informs a "Council of Agents," a novel multi-agent Large Language Model architecture where specialized agents debate and vote on a label's validity. We validate our system on a 1,000-item balanced subset of the AlleNoise benchmark. Our KG-informed model achieves a 0.99 F1-score, significantly outperforming a single-LLM baseline (0.48 F1) and a non-KG council (0.59 F1). Our analysis reveals this is due to a Precision, achieved by a novel override logic that uses the KG to perfectly identify complex, structural errors (complete Recall) -- a class of errors that baselines fail to find. This result demonstrates a robust and explainable system for automated, high-precision data verification, serving as a vital proof-of-concept for generating golden datasets in strictly governed industrial environments.