Closed
Machine Learning Text Categorization
I have created machine learning Text categorization model , during testing, sometimes when I enter random word e.g. cricket , it matches to the available topic in training data, but there is no word even close to "cricket" ( random word) in the training data, then how it is able to match it with the topic ? I want to know more about how it works ?
***Edited by Moderator Marissa to update General to Product***
Posted: 4 years ago
PEGA
Pegasystems Inc.
IN
You are probably testing with too few topics and very less training data. Take a look at few best practices documents like this to understand how to build great NLP model - https://community.pega.com/knowledgebase/articles/decision-management-overview/best-practices-creating-categorization-models