Skip to content

literature-to-table

Extracts values from papers into a table, with the source and page for every value.

  • Recipe: Literature as data and sparring partner
  • Output check: Expert-checkable A person checks a random sample of rows against the papers. Values from figures are checked more often.
  • Status: brief

Make your own version

This skill depends on local details, so the site publishes a brief instead of one institute's skill. Copy the brief and give it to your agent. If your agent has a skill-creation skill, such as Anthropic's skill-creator, it will use it. The agent asks you the questions under "Local details", builds the skill, and runs it once on the test case.

An agent that can read web pages can also fetch the brief itself from https://aiforscientists.eu/skills/literature-to-table/brief.md.

# Skill brief: literature-to-table

Source: https://aiforscientists.eu/skills/literature-to-table/
Recipe: Literature as data and sparring partner (https://aiforscientists.eu/recipes/04-literature-as-data/)

Build a skill for my own setting from this brief.
If you have a skill-creation skill or tool, use it.
Before you write anything, ask me each question under "Local details", one at a time.
Do not guess an answer; if I do not know one, leave a marked TODO in the skill.
Keep every rule under "Never" in the skill you write.
Write the check under "How to check the output" into the skill, so that it runs every time.
Then run the skill once on the test case and show me the result.

## Purpose

Extracts values from papers into a table, with the source and page for every value.

## Inputs

- The papers, with supplements.
- The columns the table needs.

## Output

- A table with one row per finding, and the source and page for every value.
- A list of papers where values could not be extracted.

## Steps

1. Read each paper and its supplements.
2. Extract one row per finding.
3. Record the source, page, and whether the value came from text, a table or a figure.
4. Mark values read from figures for extra checking.

## Local details

1. Which full-text access does your institution have, and how does the agent use it?
2. Which columns does your table need?
3. Which confounders or later corrections should the agent look for?

## Never

- Never fill in a value without a source.
- Never use full texts in a way the publisher's licence does not allow.

## How to check the output

A person checks a random sample of rows against the papers. Values from figures are checked more often.

## Test case

Ten papers you know well. Check every row.

Local details

Your version needs answers to these questions:

  1. Which full-text access does your institution have, and how does the agent use it?
  2. Which columns does your table need?
  3. Which confounders or later corrections should the agent look for?

Never

  • Never fill in a value without a source.
  • Never use full texts in a way the publisher's licence does not allow.

Local versions

No local version is published yet. If you built one, share it as an example for others.

Discussion

Used this recipe or skill, or seen it fail? Say so here. Comments need a GitHub account; without one, write to aiforscience@googlegroups.com.