Abstract

There is a great deal of interest in big data, focusing mostly on dataset size. An equally important dimension of big data is variety, where the focus is to process highly heterogeneous datasets. We describe how we use semantics to address the problem of big data variety. We also describe Karma, a system that implements our approach and show how Karma can be applied to integrate data in the cultural heritage domain. In this use case, Karma integrates data across many museums even though the datasets from different museums are highly heterogeneous.

Date

March 25, 2015

Authors

Craig A Knoblock, Pedro Szekely

Journal

Ai Magazine

Volume

36

Issue

1

Pages

25-38