In today's digital world, Ticket Management Systems are widely utilized in various businesses and organizations since they are essential for efficiently resolving client requests and difficulties. Moreover, in real-world application contexts (e.g., technology support platforms and defect detection systems), properly labeling new requests is essential to improving ticket handling quality, grade and productivity. This research aims to discover the factors affecting customers' online ticket purchasing behaviors and proposes an effective technique for systematic ticket categorization that effectively clusters and labels. Information technology (IT) tickets quickly using Machine Learning and Natural Language Processing techniques. The framework's performance in numerous ticket categorization jobs has been indicated by experimental results based on a specific usage situation that involves data from ticket mining the ServiceNow platform and filtering a vast dataset, among other things. To develop a Machine Learning model that significantly predicts an incident resolution category with supervised ML algorithms. The models' performance was estimated using various NLP techniques LDA with TF-IDF and 3 Gram, and stop-word removal and lemmatization produced the best results, with a precision of 96.46 percent.