Software has become an essential part of everyday life, and its development is producing enormous amounts of data. This includes not only source code, but also related artifacts such as change histories, test results, other execution behavior, bug reports, code reviews, and communications between developers. This constitutes an amazing wealth of rich and detailed information. At the same time, machine learning is flourishing, with a range of powerful new technologies achieving practical successes in many domains.
This workshop will bring together researchers interested in the intersection of software engineering and machine learning research. There is great promise in solving and assisting with software authoring and maintenance tasks such as coding, testing, debugging, porting, bug-patching, refactoring, optimizing, etc. Machine learning gives new tools for addressing software engineering research challenges, and software engineering challenges motivate new machine learning research. In the workshop we will discuss recent advances in this area, what challenges remain, and share ideas for how to continue progressing forward.
|9:30||Keynote: Mayur Naik|
|10:30||Contributed Talk: On Learning the Validity of Invariants|
|10:50||Contributed Talk: Improving Grey-Box Fuzzing by Modeling Program Control Flow|
|11:10||Panel Discussion: Open Research Problems in ML4SE|
|12:00||Poster Spotlights (1 min pitch per poster)|
|12:15||Lunch (on site)|
|14:00||Keynote: Cindy Rubio González|
|15:00||Contributed Talk: How Often Do Single-Statement Bugs Occur? The ManySStuBs4J Dataset|
|15:20||Contributed Talk: The Adverse Effects of Code Duplication in Machine Learning Models of Code|
|16:00||Open Discussion: Problems of Deploying ML in Software Engineering Teams|
We have reached capacity and hence cannot accomodate more registrations.
|Marc Brockschmidt||Microsoft Research|
|Prem Devanbu||University of California, Davis|
|Baishakhi Ray||Columbia University|
|Daniel Tarlow||Google Brain|