• contributors to this document

Description

CoronaWhy/task-ties

Input:

Data Dependencies

• CORD-19 Dataset
  • articles
  • metadata
  • seed articles, listed as summary-tables-2020-06-16 (located in target_tables here
  • original embeddings (from here)
    • alternate/more current(?) embeddings from CoronaWhy here (posted by Mike Honey on 8/4)

Ancillary functionalities (models generated by other notebooks?)

Currently found in CoronaWhy's Github, or the original Kaggle notebook here

FAISS engine

Time period NER model

General NER model

Sample Size NER model

Study type classifier

Output:

Tables with:

• New articles identified by similarity search (articles similar to the input seed articles)
• Extracted data (classification, NER, etc.)