The use of the semantic intelligence function.
If you have any experience with verification and experience, please let me know.
1.The result of intelligent learning shows "Error while training solution". By checking the log, it is found that the repetition rate is too high. How to create effective data
History data for training will help you. I would like to ask you how to create effective incident data
For testing Classification solutions, ideally you want to train on production data, and so you should clone/export/IDR (Instance data replication) the production data to a sub-production instance when testing your Predictive Intelligence solutions. If you create dummy data, then the trained solution would not reflect the real data in production and may produce different results. The Classifier engine will remove duplicates, so if all your dummy data contain similar text, such as "Reset my password1", "Reset my password2", etc., you may end up with a class that does not enough quality samples or is reduced to less than 30 records for the class it is trying to generate.
To generate a class, the Outfield field must have at least 30 records with the same value. Even with more than 30 records, if the data is not indicative to make a prediction on the class, it will set Precision=100, Coverage=0, Recall=0, which will set the Threshold=100 and so it will never make a prediction for this class.
