The Onion has Cancer Some Social Network Analysis Visualizations of - PowerPoint PPT Presentation

David Melissa Shapiro Marshall Thorsten Claus Barns Christian Masopust Mark S Vinod [mauj. Abhishek Parwal Stockley Tomas Kopal Maurya Ajay Henry Story Com] De Joost Greg Valk Tony Hendricks Fu Damien Miller Guzman Braso Fisher P. John Mark Mcwiggins Albert A Ting Jeffrey Vitaly Hulten Fedrushkov J. Jesse Clark gmane.comp.bug-tracking.bugzilla.devel Viagra Atsushi Pedro Sideris Stathis Shimono Lamarao Kevin Benton Manosha Emmanuel Seyman David Miller Daniel Berlin Translation India Tiago Frederic Buclin Jimderrico-re5jqeeqqe8avxti... Aaron Evans Micah Breitenstein Craig Baskinger Sam Jake Oyston Oliver Michael Tosh J Kanat-alexander Max Jason Pyeron Andreas Gervase Markham Amin Mina Mayr Jeff Lasslett Megh Man Clement Luis Villa Z. Chan G. Fuestie Noura Elhawary Nicolas Dave Doye Lawrence Ahdevans-re5jqeeqqe8avxtium... Paul Mads Bondo Byron Liu Xue Pse Dydensborg Nkg Gu Bhargava Vashista Steve Wendt Wayne Mery Dascalu Vlad Brandontkane-re5jqeeqqe8avx... G. Andres Aragoneses Craig Jean-marc Desperrier Susanna Wiedmann Jochen Tm-orsvbvaovxo@public. Sebenik Gage D. Gman... Gunnar Wagenknecht Pankaj Garg K Bill Schumann Marc Myk Melez Barry Beranek John (outsourcing) Jarrad John Wedoff Barrett Arthur Peshkin Joel Lee Dave Colin Joe Williss Ogilvie Fred Dev-apps-bugzilla-czylcwpzi... Olav Vitters Fergus Kristis Leeflang James Will Bradley Sullivan Makris Folk-williams Sam Guillaume Rousse Mcpetro Baetz Sam Fu Ke Weiling Lisa Christopher Hicks Hendricks Gregary Teemu Trevena Aaron Henry Mannermaa Nick Barnes Gupta Raj Julien Beti The Onion has Cancer Some Social Network Analysis Visualizations of Open Source Project Communication Christopher Oezbek Lutz Prechelt Florian Thiel Freie Universität Berlin ICSE Workshop on Emerging Trends in FLOSS Research and Development, 2010-05-08

2010-05-08 The Onion has Cancer David Christian Masopust Claus Thorsten Shapiro Melissa Mark Marshall Barns Henry Story Greg [mauj. Vinod S Com] Abhishek Parwal Stockley Kopal Tomas Joost Valk De Ajay Maurya Tony Miller Damien P. Fisher John Hendricks Braso Guzman Fu Mark Albert Ting A Mcwiggins Clark J. Jesse gmane.comp.bug-tracking.bugzilla.devel Jeffrey Hulten Vitaly Fedrushkov Viagra Lamarao Pedro Kevin Stathis Sideris Atsushi Shimono Manosha Translation India Tiago Seyman Emmanuel Benton Miller David Daniel Berlin Jimderrico-re5jqeeqqe8avxti... Buclin Frederic Evans Aaron Micah Breitenstein Craig Sam Baskinger Jake Oyston Oliver Michael Tosh J Max Kanat-alexander Jason Pyeron Lasslett Jeff Andreas Amin Mina Mayr Clement Villa Luis Markham Gervase Megh Man Z. G. Fuestie Dave Chan Elhawary Noura Doye Nicolas Paul Dydensborg Bondo Lawrence Mads Ahdevans-re5jqeeqqe8avxtium... Byron Steve Wendt Pse Gu Nkg Liu Xue Bhargava Vashista Brandontkane-re5jqeeqqe8avx... Craig Wayne Mery Dascalu Vlad Andres Aragoneses G. Desperrier Jean-marc Susanna Jochen Wiedmann Tm-orsvbvaovxo@public. Gman... Sebenik Gunnar Wagenknecht Gage D. Wedoff John Marc Schumann Arthur Barrett Pankaj K Garg Dave Melez Myk Barry Bill Colin Joel Peshkin Beranek John (outsourcing) Jarrad Lee Joe Will James Olav Vitters Williss Baetz Bradley Sam Sullivan Fergus Ogilvie Kristis Makris Sam Folk-williams Leeflang Fred Guillaume Rousse Mcpetro Dev-apps-bugzilla-czylcwpzi... Hendricks Gregary Mannermaa Teemu Fu Aaron Trevena Lisa Henry Christopher Hicks Weiling Ke Gupta Raj Barnes Nick Julien Beti The Onion has Cancer Some Social Network Analysis Visualizations of Open Source Project Communication Christopher Oezbek Lutz Prechelt Florian Thiel Freie Universität Berlin ICSE Workshop on Emerging Trends in FLOSS Research and Development, 2010-05-08 I would like to welcome everybody back from Lunch to this talk, titled “The Onion has Cancer. Some Social Network Analysis Visualizations of Open Source Project Communication”. My name is Christopher Oezbek and this is work done with Florian Thiel and Lutz Prechelt, who is in the audience over there, at Freie Universität Berlin.

Background ◮ The software engineering group at Freie Universität Berlin investigates software development processes and means of improving them ◮ In the context of Open Source: ◮ How can software engineering innovations be introduced into Open Source projects? ◮ Methods: ◮ Action Research ◮ Grounded Theory Methodology on mailing-list data ◮ Using theories and models from social and organizational sciences: ◮ Actor-Network Theory, Garbage Can Model, Path Dependence, Social Network Analysis , Oezbek, Prechelt, Thiel, The Onion has Cancer, FLOSS-3 2010 2

2010-05-08 The Onion has Cancer Background ◮ The software engineering group at Freie Universität Berlin investigates software development processes Introduction and means of improving them ◮ In the context of Open Source: ◮ How can software engineering innovations be introduced into Open Source projects? ◮ Methods: ◮ Action Research Background ◮ Grounded Theory Methodology on mailing-list data ◮ Using theories and models from social and organizational sciences: ◮ Actor-Network Theory, Garbage Can Model, Path Dependence, Social Network Analysis First, I want to give you some background about this work. This work is part of the software engineering group’s effort at Freie Universität to analyze software processes and determine and implement improvements. For instance, we have in recent years looked at pair programming as an important practice from XP and helped a company to achieve real-time collaboration in Eclipse between a European development center and one in india using our tool Saros. In the domain of Open Source software development—which my PhD thesis work is about—we have concentrated on understanding how software engineering innovations can be introduced into Open Source projects. This was driven by the desire to explain how volunteer organizations can achieve process change. As methods to achieve this we first used Action research—directly interacting with Open Source projects in five case studies—then we used Grounded Theory methodology on mailing-list data—as a passive and qualitative method to extract insights from past innovation introductions—and last we used several theories and models from the organizational sciences to raise theoretical sensitivity. In this last category was also Social Network Analysis, which provides the starting point for this paper.

Background ◮ The software engineering group at Freie Universität Berlin investigates software development processes and means of improving them ◮ In the context of Open Source: ◮ How can software engineering innovations be introduced into Open Source projects? ◮ Methods: ◮ Action Research ◮ Grounded Theory Methodology on mailing-list data ◮ Using theories and models from social and organizational sciences: ◮ Actor-Network Theory, Garbage Can Model, Path Dependence, Social Network Analysis , Oezbek, Prechelt, Thiel, The Onion has Cancer, FLOSS-3 2010 2

SNA and Open Source ◮ Social networks created from e-mail communication data , Oezbek, Prechelt, Thiel, The Onion has Cancer, FLOSS-3 2010 3

2010-05-08 The Onion has Cancer SNA and Open Source ◮ Social networks created from e-mail communication data Main SNA and Open Source There are many possible ways to create a Social Network can be created. For instance Greg Madey et al. used SourceForge.data to created edges between two developers if ❼ those developers worked in the same project. James Howison and Kevin Crowston created edges if people made comments on the same entry in the ❼ bug-tracker. Fernandez et al. used commits to the same module in the source code repository of a project. ❼ In this study we created the social network from the public mailing-list communication data and specifically the reply-to relationship to create edges between mailing-list participants

SNA and Open Source ◮ Social networks created from e-mail communication data A A B B A C C , Oezbek, Prechelt, Thiel, The Onion has Cancer, FLOSS-3 2010 3

2010-05-08 The Onion has Cancer SNA and Open Source ◮ Social networks created from e-mail communication data Main A A B B A C SNA and Open Source C Given a mailing-list thread as shown below, where B replied to an email by A and A and C replied to this email by B we can construct the following example network by turning each author A, B, and C into a node in the social network

SNA and Open Source ◮ Social networks created from e-mail communication data A A B B A C C , Oezbek, Prechelt, Thiel, The Onion has Cancer, FLOSS-3 2010 3

2010-05-08 The Onion has Cancer SNA and Open Source ◮ Social networks created from e-mail communication data Main A A B B A C SNA and Open Source C The reply by B to the e-mail by A is translated into an edge from B to A. The others respectively.

SNA and Open Source ◮ Social networks created from e-mail communication data A A B B A C C , Oezbek, Prechelt, Thiel, The Onion has Cancer, FLOSS-3 2010 3

SNA and Open Source ◮ Social networks created from e-mail communication data A A B B A C C , Oezbek, Prechelt, Thiel, The Onion has Cancer, FLOSS-3 2010 3