Skip to content

Dispute a label #5

Description

@evidencebp

When labeling, we try to get high accuracy.
We require justification, mark uncertainty and comment on special point.
However, the accuracy of human labeling in this domain is about 95%.

Therefore, the data set probably contain labeling error.
If you find one, please open an issue with

  • Data set file
  • The sample and its label
  • The reason to the dispute

Activity

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions