NLP meets Materials Science: Quantifying the presentation of materials data in scientific literature

The recent and sudden emergence of Large Language Models have profoundly changed the landscape around how we approach and interact with information. Materials Science, given its highly complex and multifaceted nature, is a space we intend for Natural Language Processing to absolutely flip the script related to progress, learning, and especially new materials discovery, as a result of enhanced data accessibility. We explore the underlying patterns and structures of data expression across a number of randomly selected materials science papers, annotating relevant data by type (category) and source (channel) as a starting point to future Materials Science specific information extraction and LLM development.