Labels File

The labels file is a CSV or Excel file, and contains the labels for the data file.

  • The labels file must include an ID column.
  • The labels file must include a column of labels for each variable in our dataset. The labels represent the “ground truth” as determined by a gold standard (e.g., a physician or a chart abstractor).
  • You may have multiple labels file (i.e., for training and validation).

An example labels file can be found on the Downloads page.

In the example, the ID column is used to store the IDs, and the Label column contains the labels. The name of the columns do not matter. When uploading your labels file(s) through CHARTextract’s Project settings page, you will specify the column numbers of the ID column and of the labels column.