Citation

Che Yayah, Fauzy and Ghauth, Khairil Imran and Ting, Choo Yee (2022) The Performance of Classification Method in Telco Customer Trouble Ticket Dataset. IAENG International Journal of Computer Science, 49 (2). pp. 1-10. ISSN 1819-9224

Text 15.pdf - Published Version Restricted to Repository staff only Download (1MB)

Abstract

A customer trouble ticketing system (CTT) is an organization’s tool to track the detection, reporting, and resolution of tickets submitted by customers. It also comprises a summary of the issue reported, the status of the ticket, the incident information, and the approach that was previously utilized to resolve the problems. The technician’s skill set and experience rely solely on completing the task without the right direction on which area to focus on first. As a result of this manual classification of a trouble ticket, it will be necessary to build methodologies for predicting future resolution codes. The research for this report is mainly focused on one of the telco companies in Malaysia. This study result assists the telco engineer, and the specialists resolve each issue in a very short amount of time. Additionally, the classification of the trouble ticket resolution code method used in this study will indicate the characteristics of each issue that is being investigated. The relationship between events is feasible to discover by exploring the root cause. It is critical to establish a link between recent events and events in the previous. Because of current data mining limitations, the study needs to be more comprehensive. Data processing methods are being implemented within big data platforms to overcome the limitation of data scalability, enhance classification accuracy, and increase computation speed. The research work will continue to progress in the direction of big data centricity. Some of the most effective approaches for big data integration and machine learning will be discussed in this paper. Throughout the experiment, any problems will be explained, as well as the solutions to each situation. A wide range of research subjects will be discussed, including construction classification models for trouble tickets. To achieve reasonable accuracy, a few customized transformations are required. The data set’s custom parameter optimization process will further increase the classification trouble ticket’s efficiency. However, greater processing capacity is necessitated to use multiple parallel classifiers such as Bayes, Decision-Tree, and Rule-Based with help of bigdata framewrks such as Spark. According to the study, an increase of 8% classification performance substantially influences service recovery time, customer satisfaction, and preventative maintenance expenses in the telco industry.

Item Type:	Article
Uncontrolled Keywords:	Machine learning, trouble Tickets, Sublanguage, Classification, Single Machine, Hadoop , Spark
Subjects:	Q Science > Q Science (General) > Q300-390 Cybernetics
Divisions:	Faculty of Computing and Informatics (FCI)
Depositing User:	Ms Nurul Iqtiani Ahmad
Date Deposited:	04 Jul 2022 01:28
Last Modified:	04 Jul 2022 01:28
URII:	http://shdl.mmu.edu.my/id/eprint/10124

Downloads

Edit (login required)