Knowledge Crystallisation Supported by the KnowCat System

Nowadays, it is commonly accepted that the World-Wide Web is the most popular tool for sharing knowledge and information (Berners-Lee, 1996). However, there is a huge and growing amount of information and it is getting more and more difficult to make sense out of it. The research work presented in this paper is an effort to address this Overload Information problem (Gross, 1964). The main proposal of this work consists in managing the knowledge of a user community by means of a mechanism for knowledge construction in a distributed and incremental way. More specifically, knowledge evolves towards a structured and refined state by means of user interactions. The aim of this mechanism is to achieve the crystallisation of user community knowledge as a result of user interactions, and without the need of an editor or manager of this task. The crystallised knowledge is the most accepted one by the community and, on the other hand, the knowledge that has not obtained enough acceptation will be likely to be removed. It is important to highlight that knowledge is constantly evolving. Even crystallised knowledge may receive interactions from the community for further improvement. The key point is the evolution and improvement of knowledge by means of user evaluation. The user community knowledge is subject to a maturation process involving two main phases. At first, due to the lack of critical mass of knowledge and interaction, a steering committee needs to be in charge of knowledge evaluation. Once enough mass is reached, knowledge crystallisation turns to be based on the evaluation performed by virtual communities of experts. Those users who have added knowledge that has been crystallised are considered as experts, that is, their work have been recognised by the rest of the community. Virtual communities of experts are constructed in terms of sub areas of knowledge community, and they are in charge of the collaborative evaluation of the knowledge of their sub areas. This is similar to the peer review mechanism. A collaborative Knowledge Management system called KnowCat has been designed and implemented ("Knowledge Catalyser"). KnowCat is based on the concept of Knowledge Crystallisation, supported by virtual communities of experts. KnowCat allows a user community to share, evaluate and structure collective knowledge. The system allows