Today we are joined by Anubhav Jain, Staff Scientist & Chemist at Lawrence Berkeley National Lab.
Anubhav leads the Hacker Materials Research Group, where his research focuses on applying computing to accelerate the process of finding new materials for functional applications. The paper we are discussing in this episode is ‘Unsupervised word embeddings capture latent knowledge from materials science literature'. With the immense amount of published scientific research out there, it can be difficult to understand how that information can be applied to future studies, let alone find a way to read it all. Anubhav breaks down the design of a system that takes the literature and uses natural language processing to analyze, synthesize and then conceptualize complex material science concepts. Additionally, he discusses how the method is shown to recommend materials for functional applications in the future - scientific literature mining at its best.
