Enhancing Incident Detection and Response in Cloud Computing by classifying Customer Support Cases

Raut, Aishwarya (2023) Enhancing Incident Detection and Response in Cloud Computing by classifying Customer Support Cases. Masters thesis, Dublin, National College of Ireland.

Preview	PDF (Master of Science) Download (6MB) | Preview
Preview	PDF (Configuration Manual) Download (2MB) | Preview

Abstract

This report delves into the pivotal role of incident detection and response systems in ensuring availability and minimizing downtime within cloud computing environments. This research focuses on enhancing incident detection capabilities through the application of machine learning models, specifically, non-negative matrix factorization (NMF) for unsupervised clustering and transformer models such as DistilBERT, XLMRoberta, and CNN+LSTM with Glove Embedding for supervised classification. The investigation revolves around automating the categorization of customer support tickets, addressing challenges through meticulous data loading, preprocessing, and exploration. NMF reveals distinctive patterns in support tickets, while transformer models undergo rigorous training, evaluation, and performance analysis. The research unfolds with a noteworthy achievement in the realm of incident detection within cloud computing environments. A standout result is the exceptional performance of the FineTune XLMRoberta Transformer, demonstrating high accuracy and robust categorization across various customer support ticket types. This outcome accentuates the significance of thoughtful model selection and fine-tuning, offering valuable insights into optimizing incident response strategies. However, the research is not without limitations, such as dependencies on the quality and diversity of the initial dataset and the need for periodic model updates. Despite these challenges, the results offer practical implications and feasibility of leveraging machine learning models for incident detection in cloud environments. The research contributes substantially to the field, providing a roadmap for selecting suitable algorithms and improving overall incident response efficiency. The success of integrating advanced machine learning models not only bridges theoretical gaps but also showcases practical implications for incident management in real-world scenarios.

Item Type:	Thesis (Masters)
Supervisors:	Name Email Jaswal, Shivani UNSPECIFIED
Subjects:	Q Science > QA Mathematics > Electronic computers. Computer science T Technology > T Technology (General) > Information Technology > Electronic computers. Computer science T Technology > T Technology (General) > Information Technology > Cloud computing Q Science > Q Science (General) > Self-organizing systems. Conscious automata > Machine learning
Divisions:	School of Computing > Master of Science in Cloud Computing
Depositing User:	Ciara O'Brien
Date Deposited:	10 Apr 2025 11:54
Last Modified:	10 Apr 2025 11:54
URI:	https://norma.ncirl.ie/id/eprint/7408

Actions (login required)

View Item