KGxBoard: Explainable and Interactive Leaderboard for Evaluation of Knowledge Graph Completion Models

TL;DR

KGxBoard is an interactive framework that enables detailed, interpretable evaluation of knowledge graph completion models by analyzing their performance on specific data subsets, revealing insights beyond average metrics.

Contribution

It introduces KGxBoard, a novel tool for fine-grained, explainable evaluation of KGC models on meaningful data subsets, addressing limitations of traditional single-score metrics.

Findings

Revealed model strengths and weaknesses on specific data aspects

Identified capabilities that are not apparent with average metrics

Enhanced understanding of model behavior and learning patterns

Abstract

Knowledge Graphs (KGs) store information in the form of (head, predicate, tail)-triples. To augment KGs with new knowledge, researchers proposed models for KG Completion (KGC) tasks such as link prediction; i.e., answering (h; p; ?) or (?; p; t) queries. Such models are usually evaluated with averaged metrics on a held-out test set. While useful for tracking progress, averaged single-score metrics cannot reveal what exactly a model has learned -- or failed to learn. To address this issue, we propose KGxBoard: an interactive framework for performing fine-grained evaluation on meaningful subsets of the data, each of which tests individual and interpretable capabilities of a KGC model. In our experiments, we highlight the findings that we discovered with the use of KGxBoard, which would have been impossible to detect with standard averaged single-score metrics.