Two minutes NLP — A simple taxonomy of Knowledge Graph research
Knowledge Representation Learning, Knowledge Acquisition, and Temporal Knowledge Graphs
This article is an excerpt of the paper A Survey on Knowledge Graphs: Representation, Acquisition and Applications.
Knowledge Graphs
A knowledge graph is a structured representation of facts, consisting of entities, relationships, and semantic descriptions. Entities can be real-world objects and abstract concepts, while relationships represent the relation between entities. It’s common for entities and relations to have properties or attributes.
Knowledge Graph vs Knowledge Base
The term of knowledge graph is synonymous with knowledge base with a minor difference. A knowledge graph can be viewed as a graph when considering its graph structure. A knowledge graph can be taken as a knowledge base when interpretation and inference over facts are the goals.
Current research in Knowledge Graphs
Recent advances in knowledge-graph-based research focus on knowledge representation learning, knowledge acquisition, temporal knowledge graphs, and a wide range of downstream knowledge-aware applications, where many recent advanced deep learning techniques are integrated.
Knowledge Representation Learning (KRL)
KRL consists of the mapping of entities and relations into low-dimensional vectors while capturing their semantic meanings. It’s a critical research issue of knowledge graph which paves the way for many knowledge acquisition tasks and downstream applications.
KRL deals with:
- Representation spaces in which the relations and entities are represented;
- Scoring functions for measuring the plausibility of factual triples;
- Encoding models (usually based on deep learning) for representing and learning relationships;
- Auxiliary information to be incorporated into the embedding methods, like properties of entities of relationships.
Knowledge Acquisition
Knowledge Acquisition deals with the creation of new entities and relations in the knowledge graph. Knowledge acquisition tasks include:
- Knowledge Graph Completion: inferring new entities and edges in knowledge graphs.
- Entity Recognition: extracting new knowledge graph entities from the text.
- Relation Extraction: extracting new knowledge graph factual triplets from the text.
Temporal Knowledge Graphs
Current knowledge graph research mostly focuses on static knowledge graphs where facts don’t change with time, while the temporal dynamics of a knowledge graph are less explored. However, temporal information is of great importance because the structured knowledge is true only within a specific period, and the evolution of facts follows a time sequence.
Recent research begins to take temporal information into knowledge representation learning and knowledge graph completion, which is termed as temporal knowledge graph in contrast to the previous static knowledge graph.