Make basic class to identify relevant pieces of text in larger corpora

[bramiozo]

pubscience.identify

Meta-requirements

• Generator
• Dynamic programming
• Out-of-core
• Yaml-config

Functional requirements

• text-search based
  • exact
  • fuzzy
  • document similarity-based
  • bi-encoder
  • cross-encoder
  • retrieval/rank (ColBERT etc.)
  • document-classification based

[bramiozo]

Update: student will make document-classifier with SoNaR and NtvG texts