A Survey on Knowledge Graphs: Representation, Acquisition and Applications

TL;DR

This survey comprehensively maps the knowledge-graph landscape, detailing four core strands: representation learning, knowledge acquisition, temporal knowledge graphs, and knowledge-aware applications. It provides fine-grained taxonomies across representation space, scoring, encoding models, and auxiliary information, and synthesizes embedding, path-based reasoning, and rule-based approaches for KG completion and relation extraction. It also outlines temporal extensions, practical applications in NLP and recommendation systems, and a curated set of datasets and libraries to catalyze future work. Overall, the paper emphasizes integrating symbolic and statistical methods, scalability, interpretability, and automatic construction as key directions for advancing knowledge graphs. The practical impact lies in guiding researchers toward cohesive frameworks and actionable datasets for building more capable, reasoning-aware AI systems.

Abstract

Human knowledge provides a formal understanding of the world. Knowledge graphs that represent structural relations between entities have become an increasingly popular research direction towards cognition and human-level intelligence. In this survey, we provide a comprehensive review of knowledge graph covering overall research topics about 1) knowledge graph representation learning, 2) knowledge acquisition and completion, 3) temporal knowledge graph, and 4) knowledge-aware applications, and summarize recent breakthroughs and perspective directions to facilitate future research. We propose a full-view categorization and new taxonomies on these topics. Knowledge graph embedding is organized from four aspects of representation space, scoring function, encoding models, and auxiliary information. For knowledge acquisition, especially knowledge graph completion, embedding methods, path inference, and logical rule reasoning, are reviewed. We further explore several emerging topics, including meta relational learning, commonsense reasoning, and temporal knowledge graphs. To facilitate future research on knowledge graphs, we also provide a curated collection of datasets and open-source libraries on different tasks. In the end, we have a thorough outlook on several promising research directions.

Figures (10)

• Figure 1: An example of knowledge base and knowledge graph.
• Figure 2: Categorization of research on knowledge graphs.
• Figure 3: An illustration of knowledge representation in different spaces.
• Figure 4: Illustrations of distance-based and similarity matching based scoring functions taking TransEbordes2013translating and DistMultyang2014embedding as examples.
• Figure 5: Illustrations of neural encoding models. (a) CNN nguyen2017novel input triples into dense layer and convolution operation to learn semantic representation, (b) GCN shang2018end acts as encoder of knowledge graphs to produce entity and relation embeddings. (c) RSN guo2019learning encodes entity-relation sequences and skips relations discriminatively. (d) Transformer-based CoKE wang2019coke encodes triples as sequences with an entity replaced by [MASK].

Theorems & Definitions (2)

• Definition 1: Ehrlinger and Wößehrlinger2016towards
• Definition 2: Wang et al.wang2017knowledge