Master Degree / Yüksek Lisans Tezleri

Permanent URI for this collectionhttps://hdl.handle.net/11147/3008

Browse

Filters

2016 - 20202

Settings

Subject: search.filters.subject.Model-based testing

Search Results

Now showing 1 - 2 of 2
  • Master Thesis
    A Model-Based Test Generation Approach for Agile Software Product Lines
    (Izmir Institute of Technology, 2020) Öztürk, Dilek; Tuğlular, Tuğkan
    Achieving fast development of good-quality software products is as important as achieving pure functionality. Qualified software development provides client satisfaction, reduces post-deployment costs and certificates the products. In addition to increasing quality, clients expect to tailor the products according to their needs and therefore, product configurability becomes more and more critical. Hence, the software manufacturing is required to adapt this configurable development process correspondingly. Software product line is a paradigm that purposes faster development of qualified software products that belongs to a particular domain. This thesis concentrates on quality assurance in software product lines and provides novel model-based approaches which are full test sequence composition and incremental test sequence composition approaches that aim to reuse existent test artefacts. Full test sequence composition approach reuses the existing test models and the test sequences are composed from scratch each time a product variant's test sequences are generated. Incremental test sequence composition approach reuses both of the test models and the existing test sequences of product variants. Whenever a product variant's test sequences are generated, existing test sequences and features which are incrementing the existing product are composed. The proposed approaches and the classical test generation of ESGs are compared, the results show that the incremental test sequence composition is the best in terms of both test set size and test generation time, the full test sequence composition is better than the single model ESG test generation in terms of test suite size but not in terms of test generation time.
  • Master Thesis
    Test Case Generation From Cause Effect Graphs
    (Izmir Institute of Technology, 2016) Kavzak Ufuktepe, Deniz; Ayav, Tolga; Ayav, Tolga
    Cause-effect graphing is a well-known requirement based testing technique. However, since it was introduced by Myers in 1979, there seems not to have been any sufficiently comprehensive studies to generate test cases from these graphs. Yet there are several methods introduced to generate test cases from Boolean expressions. This thesis proposes to convert cause-effect graphs into Boolean expressions and find out the test sets using test input generation techniques for Boolean expressions, such as MI, MAX-A, CUTPNFP, MUMCUT, Unique MC/DC and Masking MC/DC. Generated test sets are compared by using mutation analysis according to their fault detection capabilities. Myers’ original test generation technique is also implemented and included in the mutation analysis. A tool is created which allows to generate test cases by using the implemented algorithms. The tool gets a “.graphml” file representing a cause- effect graph as an input and gives the generated test set as an output. In addition, mutation analysis can be done with the implemented tool. 14 Requirements of TCAS-II are used as an experiment. Results of the mutation testing for these requirements showed that MUMCUT technique has the highest mutant detection success for all fault types. Moreover, Unique MC/DC technique has detected highest number of mutants per test case.