News |  Sitemap |  Contact
PDF Export  | 

Rolf Pawelzik: Knowledge Sharing Robustness of the ImageJ Community


FOSS communities as organizations perform in highly ambiguous environments, on individual as well as group scale.

This has been analyzed to negatively impact organizational 'efficiency', as goals, tasks, and resources are generally set/defined/allocated in an unpredictable ad-hoc manner. On the other hand, the unique openness and transparency of processes are analyzed to be strong motivational forces attracting a continuous stream of contributions from regular and occasional members, positively impacting organizational 'robustness'.

By operationalizing the performance of FOSS communities in terms of robustness, the ImageJ community is analyzed with regard to the level and evolution of it's knowledge sharing robustness. A robustness measure indicates how well a network's topology responses to two different events affecting the overall connectivity of the network: the first ('random' attack) simulates the effects of random member volatility, and the second ('targeted' attack) simulates the effects of a sudden disappearance of the most active members for the connectedness and integrity of the network.

To this end, the email archive of the ImageJ community from October 2004 to today has been crawled to carefully construct a series of knowledge exchange networks, representing the distribution of work load in knowledge sharing. The robustness measure is applied to these networks, indicating how this particular community managed to 'self-organize' it's work load in response to volatile demand and supply.

Regarding 'targeted' attacks, the results show that the structural robustness of the ImageJ community contains a rhythm with cycles of roughly 2 years length in which short phases of relatively high robustness follows longer phases of relatively low robustness.

This result indicates that the short phases of relatively high robustness are signs of an increased distribution of knowledge sharing work load, presumably triggered by significant changes introduced in new releases. A more careful examination of a relationship with email volume and release timing is added in the paper.


ImageJ Community, Knowledge Sharing, Organizational Robustness

Administrative data

Presenting author: Pawelzik, Rolf
Organisation: CRP Henri Tudor


© Luxembourg Institute of Science and Technology | Legal Notice