Identification of IT Tickets and Bugs using Text-Supervised Pedagogical Approaches

Abstract

“The business sector addresses the issue of task management in a number of different methods, all of which include some kind of triaging mechanism to accurately allocate tickets to developers. For Web and SaaS organizations that manage enormous amounts of tickets in the form of exceptions, support requests, user-reported bugs, and crash reports, automation of this activity has proved difficult and has resulted in less accurate results. In this work, several Machine Learning and Deep Learning algorithms for predicting assignees for new tickets based on previous tickets are investigated and analyzed. Investigating the utility of different machine learning (ML) approaches as a means of correctly constructing mathematical models for forecasting bugs and tickets was the primary focus of the effort. The research emphasizes the significance of the variables Term Frequency Inverse Document Frequency (TF-IDF) and relevant data word rate in order to guarantee that high accuracy is achieved by the Text-Supervised models that are used. In this study, an investigation of Text Supervised Learning Methods, including random forest, decision tree, and additional tree, will be carried out. The goal is to classify IT tickets and bugs more accurately.”