Matching Multiple Ontologies to Build a Knowledge Graph for Personalized Medicine

published: June 22, 2022,   recorded: June 2022,   views: 5

Related content

Report a problem or upload files

If you have found a problem with this lecture or would like to send us extra material, articles, exercises, etc., please use our ticket system to describe your request and upload the data.
Enter your e-mail into the 'Cc' field, and we will keep you updated with your request's status.
Lecture popularity: You need to login to cast your vote.
  Delicious Bibliography

Description

A rich biomedical knowledge graph can support the multidomain data integration necessary for the application of Artificial Intelligence models in personalised medicine. Constructing such a knowledge graph from already available biomedical ontologies relies on ontology matching, however, current ontology matching systems are geared towards the alignment of pairs of ontologies of the same domain one at a time. This approach, when applied to a multi-domain problem such as personalised medicine in an all vs. all fashion, poses scalability issues while also ignoring the particularities of the multi-domain aspect. In this work we evaluate a state-of-the-art ontology matching system, AgreementMakerLight, in the task of building a network of 28 integrated ontologies to construct a knowledge graph for Explainable AI in personalised oncology, highlighting its shortcomings. To address them, we have developed a novel holistic ontology alignment strategy building on AgreementMakerLight that clusters ontologies based on their semantic overlap measured by fast matching techniques with a high degree of confidence, and then applies more sophisticated matching techniques within each cluster. We implemented two within cluster alignment strategies, one based on pairwise alignment and another on incremental alignment. The within-cluster incremental alignment reduced alignment time by 80% when compared with within-cluster pairwise alignment, achieving 88% coverage of its mappings. Compared to an all vs. all pairwise approach, holistic approaches reduce total running time by up to 60%.

Link this page

Would you like to put a link to this lecture on your homepage?
Go ahead! Copy the HTML snippet !

Write your own review or comment:

make sure you have javascript enabled or clear this field: