Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

The Integration of Semantic and Structural Knowledge in Knowledge Graph Entity Typing

Muzhi Li, Minda Hu, Irwin King, Ho-fung Leung

TL;DR

This paper proposes a novel Semantic and Structure-aware KG Entity Typing (SSET) framework, which is composed of three modules and shows that SSET significantly outperforms existing state-of-the-art methods.

Abstract

The Knowledge Graph Entity Typing (KGET) task aims to predict missing type annotations for entities in knowledge graphs. Recent works only utilize the \textit{\textbf{structural knowledge}} in the local neighborhood of entities, disregarding \textit{\textbf{semantic knowledge}} in the textual representations of entities, relations, and types that are also crucial for type inference. Additionally, we observe that the interaction between semantic and structural knowledge can be utilized to address the false-negative problem. In this paper, we propose a novel \textbf{\underline{S}}emantic and \textbf{\underline{S}}tructure-aware KG \textbf{\underline{E}}ntity \textbf{\underline{T}}yping~{(SSET)} framework, which is composed of three modules. First, the \textit{Semantic Knowledge Encoding} module encodes factual knowledge in the KG with a Masked Entity Typing task. Then, the \textit{Structural Knowledge Aggregation} module aggregates knowledge from the multi-hop neighborhood of entities to infer missing types. Finally, the \textit{Unsupervised Type Re-ranking} module utilizes the inference results from the two models above to generate type predictions that are robust to false-negative samples. Extensive experiments show that SSET significantly outperforms existing state-of-the-art methods.

The Integration of Semantic and Structural Knowledge in Knowledge Graph Entity Typing

TL;DR

This paper proposes a novel Semantic and Structure-aware KG Entity Typing (SSET) framework, which is composed of three modules and shows that SSET significantly outperforms existing state-of-the-art methods.

Abstract

The Knowledge Graph Entity Typing (KGET) task aims to predict missing type annotations for entities in knowledge graphs. Recent works only utilize the \textit{\textbf{structural knowledge}} in the local neighborhood of entities, disregarding \textit{\textbf{semantic knowledge}} in the textual representations of entities, relations, and types that are also crucial for type inference. Additionally, we observe that the interaction between semantic and structural knowledge can be utilized to address the false-negative problem. In this paper, we propose a novel \textbf{\underline{S}}emantic and \textbf{\underline{S}}tructure-aware KG \textbf{\underline{E}}ntity \textbf{\underline{T}}yping~{(SSET)} framework, which is composed of three modules. First, the \textit{Semantic Knowledge Encoding} module encodes factual knowledge in the KG with a Masked Entity Typing task. Then, the \textit{Structural Knowledge Aggregation} module aggregates knowledge from the multi-hop neighborhood of entities to infer missing types. Finally, the \textit{Unsupervised Type Re-ranking} module utilizes the inference results from the two models above to generate type predictions that are robust to false-negative samples. Extensive experiments show that SSET significantly outperforms existing state-of-the-art methods.
Paper Structure (38 sections, 16 equations, 3 figures, 8 tables)

This paper contains 38 sections, 16 equations, 3 figures, 8 tables.

Table of Contents

  1. Introduction
  2. Related Work
  3. Embedding-based Methods.
  4. Graph Neural Network-based Methods.
  5. Transformer-based Methods.
  6. Semantic-aware Fine-grained Entity Typing.
  7. Problem Specification
  8. Method
  9. Step 1: Semantic Knowledge Encoding
  10. Tokenization
  11. Masked Entity Typing (MET) Task
  12. Step 2: Structural Knowledge Aggregation
  13. Embedding Fusion
  14. One-hop Knowledge Aggregation
  15. Multi-hop Knowledge Aggregation
  16. ...and 23 more sections

Figures (3)

  • Figure 1: A knowledge graph fragment centered on entity "Albert Einstein". We aim to infer missing types of the target entity based on the structural and textual information provided in the local subgraph.
  • Figure 2: The end-to-end architecture of the SSET, consists of three modules: Semantic Knowledge Encoding module (top), Structural Knowledge Aggregation module (mid), and Unsupervised Type Re-Ranking module (bot).
  • Figure 3: Experimental results of ablation studies with different experimental conditions.