Identifying annotators who did not pass the quality checks

[rmovva]

Hi, thank you for releasing this dataset. I'm doing some follow-up research with it, and wanted to perform the same quality filtering step that the paper mentions:

After all annotation tasks were completed, we performed a quality assessment on the raters and filtered out 19 raters due
to low quality work (e.g., raters who spent suspiciously little time in comparison to the other raters to
complete the task and raters who rated all conversations with the same label).

Which 19 raters were removed? Could you provide their rater_id fields, or the specific criteria you used to identify such raters?

Thank you!
Raj