Identifying Knowledge Anchors in a Data Graph

Abstract

The recent growth of the Web of Data has brought to the fore the need to develop intelligent means to support user exploration through big data graphs. To be effective, approaches for data graph exploration should take into account the utility from a user’s point of view. We have been investigating knowledge utility – how useful the trajectories in a data graph are for expanding users’ knowledge. Following the theory for meaningful learning, according to which new knowledge is developed starting from familiar entities (anchors) and expanding to new and unfamiliar entities, we propose here an approach to identify knowledge anchors in a data graph. Our approach is underpinned by the Cognitive Science notion of basic level objects in domain taxonomies. Several metrics for extracting knowledge anchors in a data graph, and the corresponding algorithms, are presented. The metrics performance is examined, and a hybridization approach that combines the strengths of each metric is proposed.