Transforming UML Models to and from Conceptual Graphs to Identify Missing Requirements
For a set of UML models built for a system, ensuring that each model is relatively complete with respect to the rest of the set is critical to further analysis and design. In this paper, we present a novel idea to identify requirements gap in a model by s
- PDF / 2,331,671 Bytes
- 8 Pages / 439.37 x 666.142 pts Page_size
- 26 Downloads / 170 Views
2
Computer Science Department, Midwestern State University, Wichita Falls, TX 76308, USA [email protected] Computer Science Department, University of Alabama in Huntsville, Huntsville, AL 35899, USA [email protected]
Abstract. For a set of UML models built for a system, ensuring that each model is relatively complete with respect to the rest of the set is critical to further analysis and design. In this paper, we present a novel idea to identify requirements gap in a model by synthesizing requirements from other types of models. This is accomplished by a bidirectional transformation between a set of partially complete UML models and conceptual graphs. UML models are first transformed to conceptual graphs based on a well-defined set of primitives and canonical graphs to form a centralized requirements knowledge reservoir, then inference rules are applied to unveil possible missing requirements. These identified missing requirements, when transformed back to UML, can provide prompts of requirements gap to software modelers and stimulate them to come up with more requirements in order to resolve them, thus making UML models more complete.
Keywords: Knowledge representation and reasoning acquisition · UML · Conceptual graphs
1
·
Requirements
Introduction
During requirements process, various types of UML models are built, each of which captures a specific view and all models together constitute the overall description of the system under development. This multiple-viewed requirements modeling approach relies on a set of interdependent models. However, it is difficult for a modeler to know whether a model is complete or what requirements are missing in the current version of the model. An effective way of finding more requirements for a model is by looking at other types of models that are being developed in parallel for the system under development: missing requirements of a model can be identified by synthesizing requirements from other types of models. c Springer International Publishing Switzerland 2016 O. Haemmerl´ e et al. (Eds.): ICCS 2016, LNAI 9717, pp. 72–79, 2016. DOI: 10.1007/978-3-319-40985-6 6
Bidirectional Transformation for Identifying Missing Requirements
73
Based on that idea, we propose a bidirectional transformation approach to identify missing requirements in a UML model. This approach adopts conceptual graphs (CGs) [6]. Since the models of a software system can be regarded as a collection of statements that evaluate to truth, i.e. assertional knowledge, CGs are appropriate for representing software requirements. Other research papers that have adopted CGs to represent requirements are [4,5]. Figure 1 illustrates the bidirectional transformation approach for the purpose of identifying missing requirements for UML models.
Class diagram State diagram
generate
convert
Conceptual Graphs Object diagram
Activity diagram
Sequence diagram
Use case diagram
Communication diagram
Fig. 1. Bidirectional transformation for missing requirements identification.
UML models are first converted
Data Loading...
