ACM DL

Software Engineering and Methodology (TOSEM)

Menu

Search TOSEM
enter search term and/or author name


Social Media

Open Continuous Special Section on AI and SE
 

CALL FOR PAPERS

Guest Editors:

  • Paolo Tonella, Università della Svizzera Italiana (USI), Switzerland
  • Tim Menzies, North Carolina State University, USA
  • Michael R. Lyu, Chinese University of Hong Kong

AI (Artificial Intelligence) technologies span a wide range of algorithms and approaches inspired by human intelligence. E.g., humans’ capability to learn from examples has given rise to a multitude of unsupervised and supervised machine/deep learning algorithms. AI technologies have seen an exponential growth in recent years, thanks to the success achieved in areas such as image processing, speech recognition, and natural language processing.

AI software is still software. And software needs maintenance, validation, interfacing, usability additions, etc. That is why AI software needs software engineers! And not only that, AI software offers special functions that need specialized approaches by software engineers. So what does engineering look like when applied to AI?

AI has been often adopted by SE researchers to address complex problems in the SE field. For instance, classifiers and neural networks have been used to predict the fault proneness of code entities. Clustering has been used to support program understanding and refactoring. Language models to automatically deal with the informal information contained in code comments or user reviews. Supervised and unsupervised models have been employed for log analysis and anomaly detection. A quite recent trend goes in the opposite direction: with the increasing adoption of AI technologies in production systems, their engineering and quality assurance demands for the kind of systematic approach that SE can deliver. Generation of adversarial test inputs, definition of metamorphic oracles and categorisation of AI specific bugs, engagement of fault tolerant techniques in deep learning models for reliability and safety are just a few examples of research areas that are actively transferring SE knowledge to the AI field.

This open special section of TOSEM looks at both direction and the topics of interest include, but are not limited to:

  • Ethical implications of AI: how design factors can lead to AI software
  • that is more/less ethical
  • SE processes and metrics applied to AI system development
  • AI for fault prediction, localization and repair
  • Analysis of the natural language in SE artefacts such as codes, comments,
  • logs and user reviews
  • Quality assurance, formal verification and testing of AI based systems
  • AI to support automated software testing
  • AI for program understanding, transformation, generation, and completion
  • Architectures and models for systems that contain AI components
  • Safety and security of AI based systems

ACM TOSEM Continuous Special Sections are journal sections that span across several consecutive issues and focus on emerging hot topics. Special section papers can be submitted at any time during the submission window, are reviewed within 90 days from the submission and with an upper bound of 45 days in the case of a major revision revision. Special section papers shall be fast-impact track papers (papers that report completely new research results or presents novel contributions that significantly extend and were not previously reported in prior work with a page limit of 45 TOSEM pages not including bibliography) and in the scope of the call.  A guest editorial board takes care of managing the review process.

Important dates: the call will remain open from October 1st, 2019 to December 31st, 2020. During this period submissions can be made at any time.

Submission: https://mc.manuscriptcentral.com/tosem
Choose Special Section on AI and SE as submission type. Papers submitted to this special section are handled as fast-impact track papers. Hence, they must comply with the constraints of such track:

  • (1) they must qualify asjournal first papers;
  • (2) their length can be up to 40 pages (plus bibliography).
 
All ACM Journals | See Full Journal Index