|
We present the application of multi-objective optimisation analytic methodologies to goal models in this research, with the intention of providing various benefits beyond the initial modelling act. Optimisation analysis can be used by modellers to evaluate goal satisfaction, evaluate high-level design alternatives, aid analysts in deciding on high-level requirements and system design, verify the sanity of a model, and improve communication and learning. Goal model analysis may be done in a variety of methods, depending on the nature of the model and the study’s goal. In our work, we use the Goal-Oriented Requirement Language (GRL), which is part of the User Requirements Notation (URN), a new International Telecommunication Union (ITU) recommendation that offers the first standard goal-oriented language. Existing optimisation methods are geared towards maximising objective functions. On the other hand, real-world problems necessitate simultaneous optimisation of both maximising and minimising objective functions. This work explores a GRL model analysis that may accommodate the conflicting goals of various inter-dependent actors in a goal model using the Analytic Hierarchy Process (AHP). By evaluating the qualitative or quantitative satisfaction levels of the actors and intentional elements (e.g., objectives and tasks) that make up the model, we construct a multi-objective optimisation method for analysis using the GRL model. The proposed hybrid technique evaluates the contribution of alternatives to the accomplishment of top softgoals. It is then integrated with the top softgoals’ normalised relative priority values. The integration result may be utilised to assess multiple alternatives based on the requirements problem. Although the URN standard does not mandate a specific propagation algorithm, it does outline certain criteria for developing evaluation mechanisms. Case studies were used to assess the viability of the suggested approach in a simulated environment using JAVA Eclipse and IBM Cplex. The findings revealed that the proposed method can be used to analyse goals in goal models with opposing multi-objective functions.
|
