Forecasting failure of information technology projects using an adaptive neuro-fuzzy inference system
Date of Award
Master of Business by Research
School of Business and Law
Dr Reza Kiani Mavi
Associate Professor Ferry Jie
Field of Research Code
The role of information technology (IT) applications has become critical for organisations in various sectors such as education, health, finance, logistics, manufacturing and project management. IT applications provide many advantages at strategic, management and operational levels, and the investment in IT applications is therefore growing; however, the failure rate of IT projects is still high, despite the development of theories, methodologies and frameworks for IT project management in recent decades. The consequences of failure of an IT project can be devastating, and can threaten the existence of an organisation. There are many different factors that impact on the performance of a project; these factors are varied and interrelated, and can impact project performance throughout the different phases of the project life cycle.
The aims of this research are to (i) identify the critical failure factors (CFFs) of IT projects; (ii) categorise these CFFs; (iii) identify the relationships between CFFs; and (iv) develop a model using an adaptive neuro-fuzzy inference system (ANFIS) to forecast the failure of IT projects in the early stages. The primary data collection tool is a questionnaire, and the analysis is carried out with the ANFIS technique.
ANFIS is a hybrid model that combines an artificial neural network (ANN) with learning algorithms and techniques, and uses fuzzy logic to extract fuzzy rules based on prior knowledge of past data. In this research, we develop 266 rules and then test the performance of the developed model using training data and checking data. In this way, the role structure of the ANFIS model is obtained, which can be used to forecast the failure of IT projects.
The findings suggest that there are many failure factors that can impact negatively on the performance of IT projects. These factors can be categorised into organisational, project management, planning, project manager, project team, user/customer, technological and technical, and legal factors. The results show that CFFs related to the project team, planning and organisation have the highest impact on the failure of IT projects.
The ANFIS model constructed here can help IT project managers to effectively address the risk associated with projects in the early phases and to forecast the failure percentage of IT projects. This research can enable managers and decision makers to predict failure early in the project, allowing them to take suitable decisions, and can provide policy makers with an innovative approach to enhance decision-making processes
This thesis is embargoed and will be unavailable until 26 November 2021.
Appendix A and Appendix B are not available in this version of the thesis.
Access to this thesis is restricted to current ECU staff and students. Email request to email@example.com
Khanfar, A. A. (2019). Forecasting failure of information technology projects using an adaptive neuro-fuzzy inference system. https://ro.ecu.edu.au/theses/2262