Customer helpdesk system plays an important role in assisting the end users or customers of the organization to get the resolutions for their service-related problems. In a typical customer helpdesk service environment, manual classification of tickets may involve misclassification and hence results in addressing the ticket to a wrong domain expert group. There is a need to develop an automated ticket classifier system which does the auto categorization of helpdesk tickets. This research paper presents such an automated helpdesk ticket classifier by using the artificial intelligence concepts like text document classification and natural language processing techniques. The proposed helpdesk ticket classifier model categorizes the incoming ticket by mining the unstructured text description entered by the end user. The research work uses the vector space model with TF-IDF term weighting approach for the representation of helpdesk tickets and Chi-square term selection technique for the dimensionality reduction. Finally, the classification techniques like linear Support vector machines, ID3 Decision trees and ensemble Random Forest are used to build an automated ticket classifier model. Real world helpdesk ticket datasets belonging to two different domain areas are used for the experimental purposes. The effectiveness of the chosen ticket classifier models is measured using various model evaluation metrics. Ensemble based Random Forest classifier performed well when compared to all other considered models. Automated ticket classifier systems result in faster ticket resolution, effective resource utilization and enhanced growth in business.

Text mining Helpdesk systems Ticket classifier Feature selection Support vector machines Random Forest