|Jan 22 – Giving a keynote at LASD22 on agility in ML engineering.|
|Jan 17 – Confirmed covid-19 in the household. We’re ok, but things are a bit messy at home.|
|Jan 10 – Back at work. Preparing for several paper deadlines next week.|
Dr. Markus Borg is a senior researcher at the intersection of software engineering and applied artificial intelligence. He is part of the Humanized Autonomy unit at RISE Research Institutes of Sweden and an adjunct lecturer at Lund University. He is also a board member of the trade organization Swedsoft and he serves on the editorial board of Empirical Software Engineering. Contact him at firstname.lastname@example.org or @mrksbrg
The goal of my research is to support the successful engineering of software and data-intensive systems. My main contributions deal with tapping into the collected wisdom of historical project data to facilitate machine learning for actionable decision support in defect management, for example, bug assignment and change impact analysis. The most useful approaches are currently implemented in tools used at Ericsson.
My research interests are related to software engineering for machine learning, especially for ADAS development mandated by automotive safety standards. Ongoing studies involve requirements engineering, MLOps pipelines, and software testing in the simulator ESI Pro-SiVIC. Currently, his work is primarily funded through two research projects:
I am the PI of the project AIQ Meta-Testbed, investigating the potential to establish a testbed for testing AI testing. The primary focus is to explore novel approaches to test deep neural networks from quality perspectives such as correctness, robustness, and model fit. However, AIQ also considers the bigger picture of how quality assurance must evolve to tackle software solutions whose logic has been trained with machine learning instead of expressed in explicit source code statements. While AIQ does indeed deep dive into neural networks to test individual activations, the project also considers broader questions regarding trusted AI. Examples include fairness testing, data validation, ethics, and the EU Assessment List for Trustworthy Artificial Intelligence. AIQ Meta-Testbed is a two-year project funded by the city of Helsingborg (2020-2021).
I am a WP lead in the SMILE project, Safety analysis and verification/validation of MachIne LEarning based systems. SMILE explores the software side of increased autonomy in the automotive domain. In the near future, vehicles will rely on deep learning to implement perception. However, safe behavior cannot be assured using traditional software engineering approaches, e.g., code review and complete code coverage testing. In this series of Vinnova FFI Machine Learning projects, RISE develops a safety cage around a deep learning component in collaboration with Volvo Cars, AB Volvo, QRTECH, Semcon, Infotiv, and ESI Group. The goal of the project is to develop mechanisms that support compliance with emerging safety standards such as ISO/PAS 21448 SOTIF. The series of SMILE projects started in 2017 and will run until 2022.
Completed projects (selected)
- ITEA3 TESTOMAT (2017-2020) – EU project on the next level of test automation. Co-applicant. WP lead. RISE project manager.
- Orion (2015-2019) – Decision support for component-based software engineering of cyber-physical systems. Funded by the Knowledge Foundation.
- EASE (2010-2015) – The Industrial Excellence Centre for Embedded Applications Software Engineering. Completed my PhD thesis on alignment between requirements engineering and software testing.
M. Borg, K. Wnuk, B. Regnell, and P. Runeson, Supporting Change Impact Analysis Using a Recommendation System: An Industrial Case Study in a Safety-Critical Context, IEEE Transactions on Software Engineering, 43(7), pp. 675-700, 2017. Rigorous empirical evaluation of the RecSys ImpRec. Lab experiments combined with deployment in the field.
M. Borg and P. Runeson, Changes, Evolution and Bugs – Recommendation Systems for Issue Management, In Recommendation Systems in Software Engineering, pp. 477-509, 2014. A book chapter on bug duplicate detection and assisted change impact analysis.
M. Borg, P. Runeson, and A. Ardö, Recovering from a Decade – A Systematic Mapping of Information Retrieval Approaches to Software Traceability, Empirical Software Engineering, 19(6), pp. 1565-1616, 2014. Remains the most comprehensive overview of IR-based trace recovery.
L. Jonsson, M. Borg, D. Broman et al., Automated Bug Assignment: Ensemble-based Machine Learning in Large Scale Industrial Contexts, Empirical Software Engineering, 21(4), pp. 1533-1578, 2016. The largest study on automated bug assignment in proprietary contexts. Introduces bug ensembles using stacked generalization.
M. Borg, TuneR: A Framework for Tuning Software Engineering Tools with Hands-on Instructions in R, Journal of Software: Evolution and Process, 28(6), pp. 426-459, 2016. A tutorial paper on how to use “design of experiments” to tune parameters of software engineering tools.