Improving your Enrichment model by adding more labels
Overview
When your Enrichment model is in production there may be times when you want to improve performance, for example when the initial training was on limited examples or your data set is changing overtime making the Enrichment model less effective.
One option is to use Active Learning to review and correct generated predictions and turn those into new labels for training. A second option is to add in more labels either from the existing data source or from a new file of labels which relates to the existing data in the Entity i.e. you have a label for a previously unlabelled entry in your Entity. This support document covers this second option
ONLY CUSTOM ENRICHMENTS SUPPORTED
At this time adding new labels or correcting labels for a Graft provided Enrichment is not supported
Next steps
- Add more labels from the EXISTING data source
OR
- Add more labels from a NEW data source