| Abstract |
Web scale reasoning has become a crucial issue for practical
applications of the Semantic Web, because of extremely large scale
data on the Web. In this talk, we propose an approach of granular
reasoning in which variable precisions/perspectives of large scale
data can be selected for reasoning to solve the scalability problem.
Several strategies under the notion of granular reasoning are
investigated, which include multiview reasoning, multilevel reasoning,
reasoning with a starting point. We investigate the granular reasoning
with a case study using an linked open dataset "SwetoDBLP". The
experimental results shows that granular reasoning is a promising
approach for Web scale reasoning.
(This talk is about an in progress research which is a joint work
between me and Zhisheng Huang.)
|