Scaling the data mountain with ontologies, deep learning and FAIR [SciBite webinar]

The explosion of data in the life sciences has led many organisations to review their approaches to good data stewardship and integration. Such concerns have given rise to initiatives such as FAIR (Findable, Accessible, Interoperable, and Reusable) which aim to promote integrity and re-use of data. However, as over 80% of organisational knowledge is stored in textual form (documents, protocols, experimental meta-data and so on) this must also be considered. Such content is notoriously difficult to utilise due to the lack of scientific understanding of the systems in which it is stored.

Recent developments in machine/deep learning and natural language processing have opened up exciting new avenues in data integration, relationship identification and smart-search. Algorithms such as BERT, ALBERT and XLNet offer much promise, but domain-specific tuning, development of training and test data and cross-corpus application still provide some challenges.

In this webinar we will:

• Describe SciBite’s semantics driven strategy for deep learning systems,
• Highlight advances in model tuning and building that allow deep learning based systems to leverage the power of life-science ontologies and the synergies this approach brings,
• Highlight a number of application areas – including the creation of vocabularies and ontologies and relationship extraction (for instance identifying protein-protein interactions).

We believe that the combination of ontologies, deep learning and FAIR data provide an unparalleled technology set that can directly impact challenges in pharma, agri-science and consumer goods. This webinar should interest anyone looking to leverage the latest deep learning techniques to bring clarity to their data.