Cognitive Knowledge Graph Reasoning for One-shot Relational Learning

TL;DR

CogKR is a novel cognitive-inspired model for one-shot knowledge graph reasoning that effectively infers new relations from minimal data by combining evidence collection and relational reasoning within a cognitive graph.

Contribution

The paper introduces CogKR, a new one-shot KG reasoning framework inspired by cognitive science, integrating evidence gathering and reasoning with explainability.

Findings

Significantly outperforms previous models on one-shot KG benchmarks.

Achieves 24.3%-29.7% relative improvement in MRR.

Provides explainable reasoning processes through cognitive graphs.

Abstract

Inferring new facts from existing knowledge graphs (KG) with explainable reasoning processes is a significant problem and has received much attention recently. However, few studies have focused on relation types unseen in the original KG, given only one or a few instances for training. To bridge this gap, we propose CogKR for one-shot KG reasoning. The one-shot relational learning problem is tackled through two modules: the summary module summarizes the underlying relationship of the given instances, based on which the reasoning module infers the correct answers. Motivated by the dual process theory in cognitive science, in the reasoning module, a cognitive graph is built by iteratively coordinating retrieval (System 1, collecting relevant evidence intuitively) and reasoning (System 2, conducting relational reasoning over collected information). The structural information offered by the cognitive graph enables our model to aggregate pieces of evidence from multiple reasoning paths and explain the reasoning process graphically. Experiments show that CogKR substantially outperforms previous state-of-the-art models on one-shot KG reasoning benchmarks, with relative improvements of 24.3%-29.7% on MRR. The source code is available at https://github.com/THUDM/CogKR.