Use the data distribution analysis tool

In a Smart Ticket configuration, on the Training Optimization tab, the Source Data Coverage parameter has a default value of 90, which means 90% of the total source data records are used to build a classification system.

The default value of 90 is the recommended value, which means with 90 percent of the training samples, the system can capture the best description of your data to build the intelligence, which will be used to categorize new tickets later. However, it may not always work well for all customer scenarios, and this is where the data distribution analysis tool can help. The data distribution analysis tool can help you visualize your tickets distribution among all of your categories, and render the results in descending order.