Software Engineering Research Patterns Mining Focus Group
Software engineering research follows a variety of paradigms, including empirical studies, formal analysis, system building, systematic literature review, and qualitative studies. Unfortunately, the field has generally not been very explicit about how research in these styles should be reported, especially about the differences between reports of research in different styles. The problem lies not just in the surface qualities of the papers, but in the deep structure – the scientific arguments. Careful reading of good papers reveals considerable regularity. This calls out for a pattern language that captures the recurring elements: identifying the settings where they are appropriate, the ground rules for carrying them out, and suitable ways to use them in combination.
The bare bones of such a Pattern Language for Software Engineering Research Papers are under development at SEresearchpatterns.org. There are patterns that capture the structure of the scientific argument and the way it maps to the overall structure of the paper. There are patterns for elements within papers (e.g., traceability from claim to result, related work) that are used in numerous types of papers. To get started, it focuses on 8-10 page conference papers.
The goal of this focus group is to put some meat on those bare bones. That means identifying patterns for more paradigms, recognizing commonalities and differences, and refining the levels of granularity in the language. We will try to capture the essence of the resulting patterns but probably will not have time to write them in detail.
It would be very helpful if participants would choose three or four papers in their preferred research style and sketch the patterns that they find in the style of http://SEresearchpatterns.org.