Wikis have their seeds in the easy collaborative editing and maintenance of web pages. This was picked up by tremendously successful public projects such as the online encyclopedia Wikipedia. Creating, modifying and maintaining of wiki articles implies social structures and dependencies between wiki authors and wiki articles themselves. The general challenge of this work is to consider these structures as dynamic evolving networks and to point out prominent behaviors in large wiki-based networks. We present an environment capable of handling data management, measurement and visualization issues for the dynamic network analysis of publicly available wiki data.
Straightforward expertise localization is crucial for personal and organizational efficiency and productivity. Common issues with existing approaches are the amount of effort required to build up the database and keeping it up-to-date, and the difficulty of establishing a complete coverage of the organization. The ExpertFinder system employs the concepts of referral chaining, social networks and user-generated data to enable a fast, low-effort and thereby low-cost approach to building an expertise localization database. At the same time, ExpertFinder provides a number of views on this database. We describe the design of the system and discuss in detail two pilots conducted at Océ-Technologies and the Telematica Instituut, the findings of which show high user participation and a good coverage of individual employee expertise and overall organizational expertise.