3536460 - Ticket NLP Classification Model Training Getting Stuck at 50% and Status "In Progress".

Machine learning model for Ticket NLP Classification models is getting stuck at 50%.

• Cloud for Customer.

 

1. Go to Administrator > Prediction services > machine learning scenarios.
2. Click on Ticket NLP Classification scenario then click on add model.
3. Select the following fields: Language, Sentiment, Product ID and Customer ID.
4. Click on train.
5. After a while click on get status and confirm that it is set to 50% and no longer progressing.

 

This is most likely happening due to lack of data for products or individual customers. You can check if this is the case by trying to train a model just for language and sentiment since language and sentiment doesn't require a certain data volume. If this processes successfully this means the issue is most likely not to do with a technical issue but is is due to the lack of volume of data for products or individual customers in the system.

The Ticket Natural Language Processing Classification scenario includes the following entities:

• Ticket language - no data needed.
• Sentiment - no data needed.
• Product ID - Requires a sufficient amount of products in the system.
• Serial ID - Requires a sufficient amount of products in the system.
• Sales Order ID - Requires a sufficient amount of Sales orders in the system.
• Customer ID - Requires a sufficient amount of individual customers in the system.

To train the functionalities other than Ticket Language and Sentiment please ensure you have sufficient data volume and good data quality for the model to use in its training.

Ticket, NLP Classification, Machine Learning, Product, Customer, ID, C4C, Cloud for Customer, Individual Customer, Sentiment, Language , KBA , LOD-CRM-ML , C4C Machine Learning , Problem

SAP Cloud for Customer add-ins all versions ; SAP Cloud for Customer core applications all versions