# Path Ranking with Attention to Type Hierarchies

## TL;DR

This paper introduces Attentive Path Ranking, an attention-based RNN model that leverages type hierarchies to improve knowledge base completion by discovering more accurate and generalizable path patterns.

## Contribution

It proposes a novel attention-based path pattern representation using type hierarchies and an end-to-end trained RNN model for knowledge graph completion.

## Key findings

- Outperforms existing methods on WN18RR and FB15k-237 datasets.
- Achieves 26% and 10% significant improvements in fact prediction.
- Path patterns effectively balance generalization and discrimination.

## Abstract

The objective of the knowledge base completion problem is to infer missing information from existing facts in a knowledge base. Prior work has demonstrated the effectiveness of path-ranking based methods, which solve the problem by discovering observable patterns in knowledge graphs, consisting of nodes representing entities and edges representing relations. However, these patterns either lack accuracy because they rely solely on relations or cannot easily generalize due to the direct use of specific entity information. We introduce Attentive Path Ranking, a novel path pattern representation that leverages type hierarchies of entities to both avoid ambiguity and maintain generalization. Then, we present an end-to-end trained attention-based RNN model to discover the new path patterns from data. Experiments conducted on benchmark knowledge base completion datasets WN18RR and FB15k-237 demonstrate that the proposed model outperforms existing methods on the fact prediction task by statistically significant margins of 26% and 10%, respectively. Furthermore, quantitative and qualitative analyses show that the path patterns balance between generalization and discrimination.