Hybrid fuzzy based decision model : a case study of web development platforms selection and evaluation

• Main Author: Ramadhan, Bwambale Rashid (Author)
• Format: Thesis
• Published: 2013-11.
• Subjects: TK Electrical engineering. Electronics Nuclear engineering; Thesis
• Online Access: Get fulltext

 Software industries are resorting to the use of web platforms due to their ability to provide access to files and information locally, remotely and on mobile devices without software prerequisites. This has resulted in a multitude of software vendors and hence a large number of web development platforms with large number of conflicting merits classifying this problem among complex decision problems. Decision making frameworks that consist of models have been successfully applied to different decision problems and have delivered dependable solutions. Multi-criteria decision frameworks like Analytical Hierarchy Process (AHP) and Technique for Order Performance by Similarity Ideal Solution (TOPSIS) are usually applied in decision activities. In most complex decision problems, the two frameworks have been integrated in a fuzzy environment to form a model due to the uncertainties in data collection. The integrated AHP and TOPSIS is complex in data collection which is performed in two phases with no inconsistency measures in the second phase; this leads to unreliable and inaccurate results. To make the results more accurate and reliable, this study has reviewed, analysed the integration of AHP and TOPSIS, investigated how the deficiencies therein can be mitigated and proposed an appropriate model named Hybrid Fuzzy Based Decision Model (HFBDM). In this model, data can be collected in either crisp or fuzzy formats and is able to determine inconsistencies in the data. This feature validates and eases data collection thereby solving the complexity and hence increasing reliability and accuracy which is the novelty and the contribution of HFBDM. The model has been evaluated and applied in a case study where data were collected in crisp format and the results demonstrate that HFBDM has more accurate and reliable outcomes compared to evaluated existing frameworks