Ontology Embedding: A Survey of Methods, Applications and Resources

TL;DR

This survey formalizes ontology embedding and faithfulness, distinguishing simple, complex, literals, and KG-augmented ontologies, and defines a concise framework for embedding ontologies. It categorizes methods into three technical solutions—Geometric Modeling, Sequence Modeling, and Graph Propagation—and covers target ontologies including simple taxonomies, DL ontologies such as $EL^{++}$ and $ALC$, as well as ontologies with literals or with KG schemas. It reviews 80+ papers, analyzes embedding performance, complexity, and interpretability, and introduces mOWL, a library to implement and benchmark these methods on tasks like subsumption prediction and PPI. It also discusses knowledge engineering applications, machine learning augmentation, life sciences uses, and highlights challenges and future directions, notably integration with large language models and neural-symbolic systems.

Abstract

Ontologies are widely used for representing domain knowledge and meta data, playing an increasingly important role in Information Systems, the Semantic Web, Bioinformatics and many other domains. However, logical reasoning that ontologies can directly support are quite limited in learning, approximation and prediction. One straightforward solution is to integrate statistical analysis and machine learning. To this end, automatically learning vector representation for knowledge of an ontology i.e., ontology embedding has been widely investigated. Numerous papers have been published on ontology embedding, but a lack of systematic reviews hinders researchers from gaining a comprehensive understanding of this field. To bridge this gap, we write this survey paper, which first introduces different kinds of semantics of ontologies and formally defines ontology embedding as well as its property of faithfulness. Based on this, it systematically categorizes and analyses a relatively complete set of over 80 papers, according to the ontologies they aim at and their technical solutions including geometric modeling, sequence modeling and graph propagation. This survey also introduces the applications of ontology embedding in ontology engineering, machine learning augmentation and life sciences, presents a new library mOWL and discusses the challenges and future directions.

Figures (9)

• Figure 1: A fragment from the OWL ontology FoodOn dooley2018foodon.
• Figure 2: Demonstration of the ontology embedding solution of geometric modeling with the examples of ELBE and $\text{Box}^\text{2}\text{EL}$.
• Figure 3: Demonstration of the ontology embedding solutions of sequence modeling and graph propagation.
• Figure 4: Dimensions, their values and corresponding works of embedding simple ontology.
• Figure 5: Dimensions, their values and corresponding works of embedding complex ontology.

Theorems & Definitions (6)

• Example 1
• Definition 1: Embedding (mathematics)
• Definition 2: Embedding (machine learning)
• Definition 3: Faithfulness of embedding (machine learning)
• Definition 4: Ontology embedding
• Definition 5: Faithful ontology embedding