Logo des Repositoriums

Auflistung nach Autor:in "Herbold, Steffen"

1 - 10 von 10
  • ConferencePaper
    Are Unit and Integration Test Definitions Still Valid for Modern Java Projects? An Empirical Study on Open-Source Projects
    (Software Engineering 2021, 2021) Trautsch, Fabian; Herbold, Steffen; Grabowski, Jens
    The article "Are unit and integration test definitions still valid for modern Java projects? An empirical study on open-source projects" published in the Journal of Systems and Software in 2020 presents the results of our investigations of the defect detection capability of unit and integration tests. While the software development context evolved over time, the definitions of unit and integration tests remained unchanged. There is no empirical evidence, if these commonly used definitions still fit to modern software development. We evaluate if the existing standard definitions of unit and integration tests are still valid in modern software development context through the analysis of the defect types that are detected, because there should be differences according to the standard literature. We classify test cases according to the definition of the IEEE and use mutation testing to assess their defect detection capabilities. We could not find any evidence that one test type is more capable of detecting certain defect types than the other one. This implies that we need to reconsider the definitions of unit and integration tests and suggest that the current property-based definitions may be exchanged with usage-based definitions.
  • Zeitschriftenartikel
    Effizientere IT-Sicherheitstest mit Hilfe von Usage-based Testing
    (Softwaretechnik-Trends Band 35, Heft 1, 2015) Schneider, Martin A.; Herbold, Steffen
    IT-Sicherheitstests untersuchen Systeme auf sicherheitsrelevante Schwachstellen, indem diese ausgeführt werden. Eine inzwischen verbreitete Technik hierfür ist das sogenannte Fuzzing, bei dem die Schnittstellen eines Systems mit ungültigen Daten stimuliert werden. Diese können zufallsbasiert, mit Beschreibungen der Eingabedatenformate, beispielsweise mit Hilfe von Grammatiken, oder zusätzlich mit Verhaltensmodellen automatisiert erzeugt werden. Da der Eingaberaum für ungültige Daten riesig oder gar unendlich groß ist, stellt sich die Herausforderungen, wie man das System effizient mit den vorhandenen Ressourcen testet. Wir möchten hier eine Idee zur Kombination von Usage-Based Testing und IT-Sicherheitstesten vorstellen, die dieses Problem abmildern kann.
  • Konferenzbeitrag
    Exploring the relationship between performance metrics and cost saving potential of defect prediction models
    (Software Engineering 2023, 2023) Tunkel, Steffen; Herbold, Steffen
    We summarize the article Exploring the relationship between performance metrics and cost saving potential of defect prediction models [TH22], which was published in Empirical Software Engineering in 2022.
  • ConferencePaper
    A Longitudinal Study of Static Analysis Warning Evolution and the Effects of PMD on Software Quality in Apache Open Source Projects
    (Software Engineering 2021, 2021) Trautsch, Alexander; Herbold, Steffen; Grabowski, Jens
    This article summarizes our work originally published in the journal Empirical Software Engineering.
  • ConferencePaper
    On the Cost and Profit of Software Defect Prediction
    (Software Engineering 2021, 2021) Herbold, Steffen
    The article "On the cost and profit of software defect prediction" published in the IEEE Transactions on Software Engineering in 2019 propose a cost model to enable the estimation of the expected profit when using machine learning models for defect prediction. Defect prediction can be a powerful tool to guide the use of quality assurance resources. However, while lots of research covered methods for defect prediction as well as methodological aspects of defect prediction research, the actual cost saving potential of defect prediction is still unclear. We close this research gap and formulate a cost model for software defect prediction. We derive mathematically provable boundary conditions that must be fulfilled by defect prediction models such that there is a positive profit when the defect prediction model is used. Our cost model includes aspects like the costs for quality assurance, the costs of post-release defects, the possibility that quality assurance fails to reveal predicted defects. Our results show that the unrealistic assumption that defects only affect a single software artifact leads to inaccurate cost estimations. Moreover, the results indicate that thresholds for machine learning metrics are also not suited to define success criteria for software defect prediction.
  • ConferencePaper
    On the Feasibility of Automated Prediction of Bug and Non-Bug Issues
    (Software Engineering 2021, 2021) Herbold, Steffen; Trautsch, Alexander; Trautsch, Fabian
    The article "On the feasibility of automated prediction of bug and non-bug issues" published in Empirical Software Engineering in 2020 considers the application of machine learning for the automated classification of issue types, e.g., for research purposes or as a recommendation system. Issue tracking systems are used to track and describe tasks in the development process, e.g., requested feature improvements or reported bugs. However, past research has shown that the reported issue types often do not match the description of the issue. Within our work, we evaluate the state of the art of issue type prediction system can accurately identify bugs. We also investigate if manually specified knowledge can improve such systems. While we found that manually specified knowledge about contents is not useful, respecting structural aspects can be valuable. Our experiments show that issue type prediction system can be trained based on large amounts of unvalidated data and still be sufficiently accurate to be useful. Overall, the misclassifications of the automated system are comparable to the misclassifications made by developers.
  • Konferenzbeitrag
    On the validity of pre-trained transformers for natural language processing in the software engineering domain
    (Software Engineering 2023, 2023) von der Mosel, Julian; Trautsch, Alexander; Herbold, Steffen
    We summarize the article On the validity of pre-trained transformers for natural language processing in the software engineering domain [VTH22], which was published in the IEEE Transactions on Software Engineering in 2022.
  • Konferenzbeitrag
    Problems with with SZZ and Features: An empirical assessment of the state of practice of defect prediction data collection
    (Software Engineering 2023, 2023) Herbold, Steffen; Trautsch, Alexander; Trautsch, Fabian; Ledel, Benjamin
    We summarize the article Problems with with SZZ and Features: An empirical assessment of the state of practice of defect prediction data collection [He22], which was published in Empirical Software Engineering in 2022.
  • Konferenzbeitrag
    Smoke testing for machine learning: simple tests to discover severe bugs
    (Software Engineering 2023, 2023) Herbold, Steffen; Haar, Tobias
    We summarize the article Smoke testing for machine learning: simple tests to discover severe bugs [HH22], which was published in Empirical Software Engineering in 2022.
  • ConferencePaper
    A Systematic Mapping Study of Developer Social Network Research
    (Software Engineering 2021, 2021) Herbold, Steffen; Amirfallah, Aynur; Trautsch, Fabian; Grabowski, Jens
    The article "A systematic mapping study of developer social network research" published in the Journal of Systems and Software in 2020 presents the results of a systematic mapping study of the state of the art of developer social network research. Developer social networks (DSNs) are a tool for the analysis of community structures and collaborations between developers in software projects and software ecosystems. We identified 255 primary studies on DSNs. We mapped the primary studies to research directions, collected information about the data sources and the size of the studies, and conducted a bibliometric assessment. We found that nearly half of the research investigates the structure of developer communities. Other frequent topics are prediction systems build using DSNs, collaboration behavior between developers, and the roles of developers. Moreover, we determined that many publications use a small sample size regarding the number of projects, which could be problematic for the external validity of the research. Our study uncovered several open issues in the state of the art, e.g., studying inter-company collaborations, using multiple information sources for DSN research, as well as general lack of reporting guidelines or replication studies.