Using multi-coordinated views with agent communication protocol to detect and resolve inconsistent requirements to improve accuracy

There are few challenges with eliciting requirements from stakeholders. Firstly, inconsistent requirements are usually not realized by stakeholders at the point of requirement specification compilation. Secondly, existing inconsistency checking techniques are usually mathematically or formally validated in a non-collaborative environment which is less practical and not productive for multiple stakeholders. In this research, a proposed requirements conflict detection and resolution solution is provided. The requirements from multiple stakeholders are captured as views on the functional requirements of a business domain using a proposed Semi-formal Ontology Driven Domain-Specific Requirements Specification Language. The proposed language will capture other relevant requirements includes domain, actors, relationships, processes, condition, and data fields to enable consistency checking. The proposed solution enables proper consistency checking in two ways: (1) Inter-Check (check within captured ontology), and (2) Intra-Check (check among stakeholders' views) so that stakeholders can resolve detected conflicts collaboratively through Agent Communication Protocol. Lab test will be conducted with a selected group of novice users to measure the learning curve (hours) for adopting the proposed solution as well as the accuracy and the consistency. A few set of requirements will be given to the group of selected users and a prototype for proposed solution will be produced. The values of this research are: (1) to improve the understanding of how much accuracy the inconsistent requirements detection is helps stakeholders and requirements analysts improve their effort in requirements conflicts detection, (2) to improve the understanding of what inconsistent requirements are detected during elicitation helps stakeholders improve the awareness of requirements conflicts.