A Comparative Study Of Sustainability Of SOURCEFORGE.net Open Source Community Projects

Internet has brought a new type of large-scale online collaborative production, the core of this model is that the community-based production behavior. Open source community is the representative. The model of open source community development is people from various places around the world to develop in collaboration, committing labor, participate spontaneously. The internal mechanism of open source community is self-organization, so the open source community can be considered as a self-organized system. How to attract users outside the community to join the community to maintain the sustainability is a major problem for open source community projects is quiet a question. Therefore, this paper intends to examine the sustainability of the open source community projects.This paper follows the basic questions, theoretical research, data extraction technology for data gathering and do empirical research. First, continuously research on open source community research literature, especially on open source community projects sustainability research. Then, to follow the trends of the internet and big data technology, extracting indicators selected. The indicators include the lifetime, the number of participants and downloaded, which on base that, data cleaning and statistical analysis according to the site classification and quantitative custom classification is followed up. Finally, from three aspects analysis of lifetime, the number of participants and downloads, we come to comparative study on sustainability of different open source community projects.There are two valuable theoretical points: 1) open source community projects are self-organized, the production is to rely on the participants in the community on a voluntary basis without any premise of reward to keep its sustainable development. There are barely related research literature on the on these new forms of self-organization sustainability. Therefore, we choose Sourceforge.net open source community projects for our study. 2) We breakthrough the limitation of locoyspider software itself to find a new way, in which we created the routine by twice data processing to complete the extraction. This kind of data extraction method has reference value for the demand of similar mining for other authors.