Multilingual Web Documents: the system Hyperling

Hyperling is a formal, language independent, system dealing with hyperdocuments (Web sites). It observes that links structure and context embed crucial information for both hyperdocument retrieving and hyperdocument mining process. For this we suggest a clustering Hyperling that deals with multilingual hyperdocuments (Web sites). In order to determine the number and frontiers between the different used languages, we adopt a distributional approach to pre process the hyperdocument structure before clustering it. Our main hypothesis considers links related to the same language be regrouped together in a cluster. From this we can conclude that the more important generated clusters represent the dominant languages