Earlier this year we found, much to our delight, that we have been granted solid research funding for our chemistry project from the Norwegian Research Council in the “BIA” grant scheme.
Since the official start of the project, we have done the necessary preparations like a couple of key hires and assigning the right researchers to the project, we have done a substantial literature review for the first work packages as well as identified the right external collaboration partners for the project.
We will keep some of the more exciting details to ourselves for now – stay in touch for more – but we can tell you as much that until spring 2022 we get to work on domain aware word embeddings models for the chemistry domain as well as chemistry embeddings models from ontologies, knowledge graph creations and graph representation models, reinforcement learning in between the knowledge graph and the chemistry embeddings.
The research project goal is partly to publish research (open access, of course!) and partly to develop software, including a commercial prototype of the “Invent” tool, which is… sort of exactly what it sounds like. The technology developed here will be an important next step for our already powerful tools for industrial R&D; navigating, identifying, extracting, and finding connections in the vast world of scientific and technical knowledge.
It has been a lot of fun getting this project started, and now that we really are digging deep into the research tasks, we are more than excited for the next 18 months.
Stay tuned for more updates on our literature review results!
(and in case you wonder – yes, of course we use our own literature review tools)