The first paper is on the new tools and artifacts used to study software development empirically, while the second paper describes how to use text mining techniques on the StackOverflow software development website.
Both works were presented at the 47th Hawai’i International Conference on System Sciences (HICSS) held on the Big Island of Hawaii from Jan. 6-9, 2014.
In the first paper, Squire gives an overview of three new artifacts that emprical software engineering researchers can use to study how free, libre, and open source software (FLOSS) is made. The main findings were that Github, Stack Overflow, and paste sites (like Pastebin, Github gist, and JSFiddle) are critical new tools for the development of FLOSS, and the artifacts from these tools can be collected and studied using similar techniques to what we have done in the past.
In the second paper, Squire and Funkhouser proposed a novel way of mining the text data from the StackOverflow developer question-and-answer web site. They were interested in determining programmatically which text metrics will lead to higher quality postings. The main findings were that a higher code-to-text ratio matters in formulating answers more than in formulating questions, and that readability does not matter as much as sentiment in constructing high-quality postings.
The full citations are below:
Squire, M. (2014). Forge++: The Changing Landscape of FLOSS Development. In Proceedings of the 47th Hawaii International Conference on System Sciences (HICSS47). IEEE. Big Island, HI, USA. 3266-3275. DOI: 10.1109/HICSS.2014.405
Squire, M. and Funkhouser, C. (2014). “A Bit of Code”: How the Stack Overflow Community Creates Quality Postings. In Proceedings of the 47th Hawaii International Conference on System Sciences (HICSS47). IEEE. Big Island, HI, USA. 1425-1434. DOI: 10.1109/HICSS.2014.185