Workshop on AI and Software Testing/AnalysisAISTA 2021
Artificial Intelligence (AI) has achieved substantial success in enhancing various software testing and program analysis techniques and applications, including but not limited to static analysis, fuzz testing, GUI testing, vulnerability detection, code similarity analysis, software debloating, and patching. We often see a synergistic effect that AI models, by learning from past experience to make decisions, can notably boost conventional program analysis and software testing tasks. Hence, it is a promising direction by applying advanced machine learning techniques into suitable software engineering tasks.
Furthermore, recent years have also witnessed a substantial adoption of AI models in safety- and security-critical applications such as medical image processing, autonomous driving, aircraft control systems, machine translation, and surveillance cameras. Thus, it is also highly crucial to apply software testing and program analysis techniques to ensure the robustness, fairness, explainability, and reliability of AI models, especially when AI is applied into safety- and security-critical applications.
The AISTA workshop aims to create an opportunity for the researchers to discuss their research, share recent ideas, and present new perspectives at the intersection of AI and Software Testing/Analysis, i.e., AI for Software Testing/Analysis and Software Testing/Analysis for AI. The workshop will consist of invited talks and presentations based on research paper submissions.
For more information please consult https://ai-sta.github.io/aista21/
Mon 12 JulDisplayed time zone: Brussels, Copenhagen, Madrid, Paris change
09:00 - 11:50
|Towards Automated Debugging: A Trace Travelling Oriented and AI-based Approach|
Yun Lin National University of Singapore
|NerdBug: Automated Bug Detection in Neural Networks|
|Automated Cell Header Generator for Jupyter Notebooks|
|Impact of Programming Languages on Machine Learning Bugs|
Sebastian Sztwiertnia Technical University of Darmstadt, Maximilian Grübel Technical University of Darmstadt, Amine Chouchane Technical University of Darmstadt, Daniel Sokolowski TU Darmstadt, Krishna Narasimhan TU Darmstadt, Mira Mezini TU Darmstadt, GermanyLink to publication DOI Pre-print
|On the use of Evolutionary Algorithms for Test Case Prioritization in Regression Testing considering Requirements Dependencies|
Call for Papers
AISTA invites the following submissions:
Full papers (up to 8 pages; including references): Original, unpublished results that are related to the topics of AISTA.
New idea papers (up to 4 pages; including references): New ideas supported by initial evidence.
Fast abstract papers (up to 2 pages; including references)
We plan to organize AISTA as an one-day workshop, including a half-day keynote session and a half-day presentation of accepted papers. Note that all papers, including full papers, new idea papers, and fast abstract papers, will receive a full presentation and discussion slot.
AISTA focuses on the intersection of AI and Software Testing/Analysis. We welcome (but not limited to) the following topics:
- AI for program static and dynamic analysis
- AI for software testing
- AI for other software engineering topics like optimization or code comprehension
- Apply software testing techniques to AI models
- Apply program static or dynamic analysis techniques to AI models
- Testing/analysis AI infrastructures
- Empirical study of related topics
Papers must conform to the ACM conference format and in PDF. Papers will be reviewed in a double-blinded process. Papers can be submitted via EasyChair (https://easychair.org/conferences/?conf=aista21). In particular, the following LaTeX code should be placed at the start of the LaTeX document.
Proceedings of AISTA 2021 will be available through the ACM DL and indexed in DBLP.