Finding Communities of Practice from User Profiles Based on Folksonomies

User profiles can be used to identify persons inside a community with similar interests. Folksonomy systems allow users to individually tag the objects of a common set (e.g., web pages). In this paper, we propose to create user profiles from the data available in such folksonomy systems by letting users specify the most relevant objects in the system. Instead of using the objects directly to represent the user profile, we propose to use the tags associated with the specified objects to build the user profile. We have designed a prototype for the research domain to use such tag-based profiles in finding persons with similar interests. The combination of tag-based profiles with standard recommender system technology has resulted in a new kind of recommender system to recommend related publications, keywords, and persons. Especially the latter is useful to find persons to potentially cooperate with and to monitor the community to be able to enhance a user's current Community of Pract...

• Type: Conference
• Year: 2006
• Where: ECTEL
• Authors: Jörg Diederich, Tereza Iofciu
• Keywords: ECTEL 2006, Folksonomy Systems, Machine Learning, Similar Interests, User Profile