Ontology Learning

Ontology learning is a relatively new field that aims to automatically (or semi-automatically) learn or create ontologies (using machine learning, text mining, knowledge representation and reasoning, information retrieval and natural language processing techniques) from some text or corpus.

Ontology learning can be divided into different phases or tasks

1. the acquisition of terms that refer to specific concepts (named-entity recognition)
2. the recognition of synonyms among these terms
3. the identification of taxonomic relations (such as the "is-a" relation)
4. the establishment of non-hierarchical relations
5. the derivation of new knowledge, i.e. knowledge that is not explicitly encoded by the ontology.

See also Ontology Learning from Text: An Overview (2003) and A survey of ontology learning techniques and applications (2018) for more details.

In the paper Ontology Learning with Deep Learning: a case study on Patient Safety using PubMed (2016), the authors investigate how continuous bag-of-words (CBOW) and skip-gram (two language models based on artificial neural networks) can be used to aid ontology development for patient safety, using PubMed citations as a corpus.

Latent Dirichlet allocation (LDA) has also been used for ontology learning, for example, in the paper Terminological ontology learning and population using latent Dirichlet allocation. (2014).

Reading List

https://www.amazon.co.uk/Knowledge-Graph-Cookbook-Andreas-Blumauer/dp/3902796707

Ontology Tools & Resources

https://souslesens.github.io/souslesensVocables/