Neural Graph Reasoning: Complex Logical Query Answering Meets Graph Databases

By H. Ren et al

Read the original document by opening this link in a new tab.

Table of Contents

Contents
1 Introduction
2 Preliminaries
2.1 Types of Graphs
2.2 Basic Graph Query Answering
2.3 Basic Approximate Graph Query Answering
2.4 Graph Query Answering
2.5 Approximate Graph Query Answering
2.6 Triangular Norms and Conorms
2.7 Graph Representation Learning
3 Neural Graph Databases
3.1 Neural Graph Storage
3.2 Neural Query Engine
4 Graphs
4.1 Modality
4.2 Reasoning Domain
4.3 Background Semantics
5 Modeling
5.1 Encoder
5.2 Processor
5.3 Decoder
5.4 Computation Complexity
6 Queries
6.1 Query Operators
6.2 Query Patterns
6.3 Projected Variables
7 Datasets and Metrics
7.1 Evaluation Setup
7.2 Query Types
7.3 Training
7.4 Inference
7.5 Metrics
8 Applications
9 Summary and Future Opportunities

Summary

Complex logical query answering (CLQA) is a task of graph machine learning that involves multi-hop logical reasoning over massive graphs. In this paper, a survey of CLQA is provided with a focus on Neural Graph Databases (NGDBs) which enhance query answering in incomplete graphs. The paper introduces a framework for NGDBs including a Neural Graph Storage and a Neural Query Engine. The text discusses the challenges with traditional graph databases and the benefits of using NGDBs. Various aspects of graph databases, modeling, queries, datasets, and metrics are explored. The paper concludes with future research directions and applications of NGDBs.