Prediction of Software Faults Based on Requirements and Design Interrelationships
Ladda ner
Typ
Examensarbete för masterexamen
Master Thesis
Master Thesis
Program
Software engineering and technology (MPSOF), MSc
Publicerad
2016
Författare
Nassar, Bashar
Modellbyggare
Tidskriftstitel
ISSN
Volymtitel
Utgivare
Sammanfattning
Traceability information between different system artifacts, like the one between requirements, architectural elements and the results of test cases can be used to expose interesting relationships between the early phases of the software development process and the software faults in the end product. For instance, complex dependencies between features and software components could lead to an increased level of flaws in the code. Such patterns can be detected and visualized as early warnings to the relevant stakeholders (e.g., the architect or the project manager). Ultimately, a fully-fledged prediction model can be developed if enough historical information is available from previous software projects. In this thesis work we investigate the relationships between the system design attributes and test case results looking for fault patterns using the traceability data and design metrics. Our intention is to use these patterns to predict the system faults in early stage of the development process. The ultimate goal is to introduce a method for building a decision support system based on historic product data. The research presented in this thesis is based on a quantitative case study conducted together with our industrial partner Systemite AB, where the raw data was provided by three Swedish automotive companies. We found that design attributes, such as a number of component in port or functionality could have a strong relation with the probability of failed tests, thus they could by use as an indicator to predict faults. Those faults could be avoided during the design phase, which will lead to improve the quality and reliability of the system and reduce its cost and development time.
Beskrivning
Ämne/nyckelord
Data- och informationsvetenskap , Computer and Information Science