Loading...

Towards Decision Support Framework for Architecture Reconstruction Based on Domain Requirements by Use of Styles and Patterns

Moaven, Shahrouz | 2021

375 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 54488 (19)
  4. University: Sharif University of Technology
  5. Department: Computer Engineering
  6. Advisor(s): Habibi, Jafar
  7. Abstract:
  8. Nowadays, due to the enlargement and complexity of software systems, traditional approaches to software development are not as efficient as before, hence there is a need for new approaches in design, implementation, and maintenance of software systems and processes. Complexity and enlargement of design and initial implementation of systems and processes on one hand, and the requirements appear in the maintenance phase on the other hand, makes the dynamic design of systems and processes a very important issue. However, adding dynamism and changeability to the systems has an upper bound, and additionally, it is impossible to definitely predict trend of systems and processes in the future. This issue causes changes in the structure and behavior of systems by time, which demands unavoidable corrections and reconstructions. Software architecture reconstruction is a critical need in working systems and processes; it not only reconstructs existing structures, but also it reviews decisions about architecture, design, and implementation logic, are both totally and by detail. Software architecture reconstruction is a challenging and complex issue which demands high knowledge and expertise and must be implemented by mechanized tools and methods. The goal of this PhD proposal is developing a framework that supports all phases of software architecture reconstruction process. This framework, by aggregating and integrating existing data and processes, and by taking advantage of cognitive science rules in all steps, and by making use of embedded specific evaluation methods, makes us capable of reconstructing architecture of systems and processes in an effective and affordable way, based on requirements engineering concepts and using solutions which are based on patterns. In this framework, by taking exploiting data gathering techniques such as data mining, all data or information dependent to the system architecture is gathered by the watchdog component. This framework, like all decision support frameworks, contains a knowledge management component which has been designed for software architecture knowledge management and provides us with capability of knowledge retrieval based on the rule-base which is filled by rules both manually and automatically. This framework provides a rational and conceptual relation using a precise methodology among different influential components. Also, it makes it possible to use different perspectives, solutions, and tools (separately or in combination) in each stage of the work. Moreover, the designed framework helps to mitigate the existing weaknesses by the use of fuzzy evaluation methods and hierarchical analysis. It is worth noting that these structures and mechanisms have been designed based on styles and patterns of software architecture design with a special focus on quality attributes
  9. Keywords:
  10. Reengineering ; Requirements Engineering ; Software Architecture ; Knowledge Management ; Decision Making Support Framework ; Architecture Refactoring

 Digital Object List