Controlling open source intermediaries - a Web log mining approach

Open Source Software (OSS) has be- come a focal research issue in computer science. One interesting phenomenon in the OSS market is the evolution of dedicated intermediaries, which provide OSS-related services to the OSS community. Prevalently, OSS intermediaries fail to fulfil their financial and non-financial goals, such that many of them have vanished from the OSS market or significantly changed their busi- ness models. A major reason for this develop- ment is the lack of appropriate instruments which are able to dismantle coordination prob- lems and failures inherent to the OSS develop- ment model. This paper proposes a controlling instrument for OSS intermediaries based on a behaviouristic approach. Since OSS intermediaries typically provide internet based services, it is possible to observe actor behaviour by use of technical data acquisition instruments like server logging. Therefore, it seems reasonable to examine the concept of web log mining, which provides a framework to analyse server based log files. Fi- nally, we apply this concept to the research field of OSS intermediaries and show empirical re- sults. These results relate to the OSS intermedi- ary CampusSource, who provides e-learning software and complementary services.