Incident Ticketing Analytics On Issue Categorization Using Text Mining Techniques
Abdul November 7, 2020
American manufacturer of consumer and professional products
Problem Statement
Business wants to know the root causes of the service request ticket raised from various applications through the unified ticketing portal.
Analyze the cause and sub-issues that are leading to increase the tickets
Solution Approach
Text Processing
Building Matrix of Words
Clustering
Drill Down
Business Impact
20% issues are due to top 2 application with UI issues are most recurring one
3% Improvement in Issue identification process through automation algorithms on Issue data
Critical Success Factors
- Text column is cleaned and pre-processed using NLP techniques
- Once the clusters are identified, each cluster is drilled down to identify the issue cause attached to ticket