Grammar-based CORD-19 Approach | Kaggle

While researching the very challenging and open ended questions posed by this Kaggle task, we take an approach inspired by Qualitative Social Science. Therefore we conducted a few unstructured interviews with policy-makers, ethics counselors, and front-line medical practitioners in order to better understand what are their biggest needs and how our work could be most helpful for them. The details that experts shared with us informed the data exploration and modeling presented in this notebook.

The goal of this notebook is to link public health measures and social and ethical concerns for individual articles in the corpus. In order to extract such information, we will utilize a language grammar. In our case we've defined a grammar that aims to capture how the literature discusses factors such as the enablers, barriers, and implications of different public health measures. Our current grammar is in English but our language grammar approach could be evolved into covering other languages as well.