The Large Knowledge Collider∶ Knowledge representation at very, VERY large scale

Name: Frank van Harmelen
Organisation: Network Institute   Computer Science Department at VU University Amsterdam
Abstract:

Traditional knowledge-bases were on the order of thousands of axioms, and knowledgebases with tens of thousands of axioms were unheard of, and certainly unmanagable. The development of the Semantic Web over the past decade has changed this for ever. There are now publically available knowledgebases with many millions of axioms, and the total number of axioms on the Semantic Web is counted in the tens of billions!

To cope with this huge increase in size, we have developed the Large Knowledge Collider, a computational platform that allows us to perform inference over such huge knowledge bases, using highly parallel algorithms and hardware.