Consistency-Aware Parameter-Preserving Knowledge Editing Framework for Multi-Hop Question Answering
This addresses reliability issues in updating AI models with new knowledge for multi-hop reasoning tasks, representing an incremental improvement over existing PPKE methods.
The paper tackles the problem of inconsistent knowledge updates in Parameter-Preserving Knowledge Editing (PPKE) for multi-hop question answering, which leads to knowledge contamination and unreliable reasoning. The proposed CAPE-KG framework improves PPKE accuracy on the MQuAKE benchmark.
Parameter-Preserving Knowledge Editing (PPKE) enables updating models with new or corrected information without retraining or parameter adjustment. Recent PPKE approaches based on knowledge graphs (KG) to extend knowledge editing (KE) capabilities to multi-hop question answering (MHQA). However, these methods often lack consistency, leading to knowledge contamination, unstable updates, and retrieval behaviors that fail to reflect the intended edits. Such inconsistencies undermine the reliability of PPKE in multi-hop reasoning. We present CAPE-KG, Consistency-Aware Parameter-Preserving Editing with Knowledge Graphs, a novel consistency-aware framework for PPKE on MHQA. CAPE-KG ensures KG construction, update, and retrieval are always aligned with the requirements of the MHQA task, maintaining coherent reasoning over both unedited and edited knowledge. Extensive experiments on the MQuAKE benchmark show accuracy improvements in PPKE performance for MHQA, demonstrating the effectiveness of addressing consistency in PPKE.