You can use Active Learning to find any outliers that might exist in previously coded data. This will compare the human coded value to a machine value for responsiveness. This disagreement might be fine, but might point to documents that are coded incorrectly. It will also help you understand the process of Active Learning and how to effectively use it for new projects.
Please refer to the Quick Start guide for Active Learning for more information.
- Create a field for Active Learning Project and add two choices with one being an affirmative value.
- Create an Analytics Classification Index. All documents included in the index will be used for project.
- Create a security group.
- Use all of the steps from project prep to create your project.
- Take a sample of your coded items based on Responsives and mass update the new Active Learning field. Then, do the same for Not Responsives.
- At this point, you introduce the previously coded items to the project.
- To add items to the Active Learning project, you can mass code the field used by the project. But don’t send more than a couple of thousand to the project start.
- Sample the database and mass code over the sample.
- Go to the project and update the ranks. This will publish the data over to the Relativity fields that are useful in finding the conflicts. To see the results visually, see Active Learning Useful Dashboards.
- Search for documents coded in the reviewer field as Responsive and Classified by the system as Not Responsive.