A hybrid technique using minimal spanning tree and analytic hierarchical process to improve functional requirements prioritization
Software for large enterprises such as the Enterprise Resource Planning (ERP) is more likely to be developed by a team of software developers where the functional requirements (FRs) are distributed in parallel developers. Therefore, development of pre-requisite FRs must be carefully timed to see whi...
Saved in:
| Main Author: | |
|---|---|
| Format: | Thesis |
| Language: | English English English |
| Published: |
2021
|
| Subjects: | |
| Online Access: | http://eprints.uthm.edu.my/1804/2/MUHAMMAD%20YASEEN%20-%20declaration.pdf http://eprints.uthm.edu.my/1804/1/MUHAMMAD%20YASEEN%20-%2024p.pdf http://eprints.uthm.edu.my/1804/3/MUHAMMAD%20YASEEN%20-%20full%20text.pdf |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Software for large enterprises such as the Enterprise Resource Planning (ERP) is more likely to be developed by a team of software developers where the functional requirements (FRs) are distributed in parallel developers. Therefore, development of pre-requisite FRs must be carefully timed to see which requirement is to be implemented first by assigning priority to some FRs over others, so that FRs can be made available on time to parallel developers. Well-known prioritization technique such as the Analytic Hierarchical Process (AHP), although accurate, is not scalable for large set of FRs as in ERP due to high number of pairwise comparisons when the size of FRs is more than ten or twelve. To address this issue, this research proposes a hybrid prioritization technique of Minimal Spanning Trees (MST) and AHP called the Spanning Analytic Hierarchical Process (SAHP) for FRs prioritization by exploiting MST capability to prioritize large size software FRs with smaller pairwise comparisons but with more consistent results. Using Numerical Assignment (NA) technique, prioritized FRs from SAHP are assigned to priority groups such that top priority groups contain high priority FRs and low priority groups contain low priority FRs. Low priority group of FRs are dependent on high priority groups. As a result, within each priority group, inter-dependencies in FRs are reduced for parallel developers. Implementing high priority groups will reduce number of dependencies in FRs among the lower priority groups. The proposed technique is evaluated based on ERP case study and the results showed that SAHP reduces estimation time of parallel developers as compared to AHP and other techniques. This shows that SAHP is scalable to cater large number of pairwise comparisons for large systems like ERP. |
|---|