OWL2Vec*: embedding of OWL ontologies

Abstract

Semantic embedding of knowledge graphs has been widely studied and used for prediction and statistical analysis tasks across various domains such as Natural Language Processing and the Semantic Web. However, less attention has been paid to developing robust methods for embedding OWL (Web Ontology Language) ontologies, which contain richer semantic information than plain knowledge graphs, and have been widely adopted in domains such as bioinformatics. In this paper, we propose a random walk and word embedding based ontology embedding method named , which encodes the semantics of an OWL ontology by taking into account its graph structure, lexical information and logical constructors. Our empirical evaluation with three real world datasets suggests that benefits from these three different aspects of an ontology in class membership prediction and class subsumption prediction tasks. Furthermore, often significantly outperforms the state-of-the-art methods in our experiments.

Contextual Semantic Embeddings for Ontology Subsumption Prediction

“…This is mainly due to the superiority of contextual text embeddings by BERT, and this result is consistent with BERTMap which also uses BERT to find equivalent classes using their labels alone (He et al 2022). It is worth mentioning that the performance ranking of the baselines in this study is mostly consistent with that in (Chen et al 2021) where all the named classes are considered for the negative subsumers in testing. Second, using additional synonym properties besides rdfs:label has positive impact on FoodOn but negative impact on GO.…”

“…Some recent ontology embedding methods take lexical annotations into consideration and are applied in class subsumption prediction. For example, OWL2Vec * (Chen et al 2021) first transforms an OWL ontology into sequences and then uses Word2Vec (Mikolov et al 2013) to learn the embeddings of the classes, and finally trains a classifier for subsumption prediction using the ontology's existing subsumptions. OPA2Vec (Smaili, Gao, and Hoehndorf 2019) and Onto2Vec (Smaili, Gao, and Hoehndorf 2018) are similar ontology embedding methods that can be used in such a subsumption prediction pipeline.…”

Discovering alignment relations with Graph Convolutional Networks: A biomedical case study

“…That is why Gutiérrez-Basulto and Schockaert [16] investigate how to ensure logical consistency through geometrical constraints on embedding spaces and if classical embedding techniques respect such constraints. Similarly, OWL2Vec* [7] focus on embedding complex logical constructors as well as the graph structure and literals.…”

Prediction of Adverse Biological Effects of Chemicals Using Knowledge Graph Embeddings

“…A snapshot of TERA is available on Zenodo [52], where licenses permit. 11 PubChem and ChEMBL are not included in the snapshot due to size constraints; these can be downloaded from the National Institutes of Health 12 and European Bioinformatics Institute, 13 respectively. The subgraph of TERA used for prediction is available alongside the chemical effect prediction models in our GitHub repository.…”

“…Each result is composed by Table 5 Example triples from the TERA knowledge graph. For space reasons, we have added the full id or label for some of the entities using footnote marks where 1 inchikey:MMOXZBCLCQITDF-UHFFFAOYSA-N, 2 Pimephales, 3 Cyprinidae, 4 Headwater, 5 Benzamides, 6 Insect Repellents, 7 CHRNA3, 8 CHRNB4, 9 DETA-20, 10 DETA Epichlorohydrin, 11 Has component, 12 Triclocarban, 13 Trichlorocarbanilide-containing product, 14 Similar to, 15 3-Chloromethyl-N,N-diethylbenzamide. an endpoint, an effect, and a concentration (with a unit) at which the endpoint and effect are recorded.…”