Open Access Open Access  Restricted Access Subscription Access
Cover Image

A Preemptive Model to Enhance the Software Design through Use Case Analysis with Slicing Technique

Lalitha R, Latha B, Sumathi G

Abstract


In software design, use case diagrams play a major role since the requirements analysis, results are depicted by use case diagrams. In object oriented software engineering, the relationship between various modules also depend on class hierarchy. In such a scenario, a careful analysis on the use case diagrams will lead to a better translation of analysis modeling to design documents. Therefore, it is necessary to provide a technique to understand the various components of use case diagrams and to make suitable design based on analysis modeling. In the past, new techniques were proposed for program slicing and dynamic slicing of software architectures. Such features are useful to enhance the design metrics. In this paper, a new design methodology is introduced to improve the software design by applying slicing technique for increasing the cohesion and for reducing the coupling. The materials and methods of this work uses the analysis documents and makes decisions based on the components of use case diagrams. For this purpose, the use cases are sliced using intelligent rules and the parameters for design are identified from these slices. Moreover, these slices are used to perform effective classification in which the parameters obtained from the use case diagram are used as features for classification. In addition, the parameters are ranked using a new feature ranking algorithm based on information gain and classification is performed using Multiclass Support Vector Machines. The analysis phase is classified based on the classes and modules so that the cohesion and coupling necessary for scheduling the activities can be identified. This work uses a preemptive model in which the important modules are given priority and high priority modules can preempt the low priority modules with the approval of a decision manager. The main aim of the proposed work is that with helps the project managers by prioritizing the activities so that the activities can be scheduled optimally. Five layers are proposed in this model which are used for creating, analyzing to find the parameters, feature ranking, classification and decision making. From the results obtained from this work, it is concluded that the proposed work provides optimal scheduling.


Full Text:

PDF

Refbacks

  • There are currently no refbacks.