Skip to content

Extraction Table

The Extraction Table page at /extractions lets you run structured data extraction across multiple papers using LLM-powered templates, and then compare and export the results.

Workflow

1. Select a template

Templates define the fields to extract — for example a template might specify columns for "Sample size", "Evaluation metric", and "Baseline model". Choose a template from the template selector dropdown.

Templates are managed in Settings → System → Extraction Templates. Creating and editing templates is an admin function; using them to extract data is available to all users.

2. Select papers

Use the PaperSearchSelect control — a multi-select search field — to choose one or more papers from your library. Type a title or author name to find papers. You can select as many papers as you need.

3. Extract Selected

Click the Extract Selected button. The system submits an LLM extraction job for each selected paper using the chosen template. A progress indicator shows extraction status per paper. Depending on the number of papers and the complexity of the template, extraction may take from a few seconds to several minutes.

ExtractionDataTable

Once extraction is complete, the results are displayed in the ExtractionDataTable:

  • Rows correspond to the fields defined in the template.
  • Columns correspond to the papers you selected.
  • Each cell shows the extracted value for that field in that paper. Cells may be empty if the model could not find a relevant value.

The table helps compare how selected papers describe the same aspects of a research problem. It can support screening and evidence organization, but every extracted value should be checked against the source before use in a systematic review or meta-analysis.

Export CSV

Click Export CSV to download the current table as a comma-separated values file. The CSV includes all visible rows and columns. You can open it in any spreadsheet application for further analysis.

  • Settings — manage extraction templates (§IV System → Extraction Templates; admin required).
  • Paper Detail — run entity extraction for a single paper from the right-hand actions sidebar.
  • Knowledge Graph — visualise entities extracted from papers.