Skip to content

PaperToKeyword_spec.md

Latest commit

 

History

History
39 lines (26 loc) · 1.1 KB

PaperToKeyword_spec.md

File metadata and controls

39 lines (26 loc) · 1.1 KB

Paper to Keywords (Text to Text)

Goal

Extract relevant keywords from lengthy scientific papers to highlight the main topics and key areas of focus.

Evaluation

Keywords will be evaluated by human reviewers based on:

  • Relevance to the text.
  • Accuracy in capturing core topics and concepts.
  • Usefulness for understanding the paper's content.

Actions

getKeywords()

  • Params:

    • file (file): File that contains the full text of the paper. Max size: 5MB.
    • keywordsCount (int): Desired number of keywords. Maximum: 50.

  • Returns:

    • keywords (array): An array of extracted keywords or phrases that represent the paper's main topics.

Performance Requirements

  • Response within 60 seconds for files <2 MB.
  • At least 200 API calls per subscription per month.

Constraints

  • Extracted keywords must be non-redundant and accurately reflect the paper's content.
  • Prohibit processing of files with unsupported formats or encrypted content.
  • Ensure keywords are generated in a language consistent with the paper's text.