Toward granular knowledge analytics for data intelligence: Extracting granular entity-relationship graphs for knowledge profiling
Alexander Denzler, M. Kaufmann · 2017
This paper proposes an approach to computational knowledge analytics. Identification and measurement of knowledge — a meta-knowledge of who knows what — is a crucial foundation for decision-making. The availability of information sources and know-how is a valuable production factor. In this paper, the authors propose a framework to capture, visualize, and analyze organizational knowledge by extracting granular, hierarchical entity-relationship models from unstructured text data that are mapped to individual users and organizational units. With a granular computing approach, the framework can be applied to model, represent, and visualize knowledge profiles that show who in the organization knows what. The framework can also be applied in big data management to enhance data intelligence — the competence, knowledge, and skills necessary to analyze and utilize big data. We present a prototype implementation that assesses and harnesses data sources and metrics to deliver a meaningful analysis of user and community knowledge, based on text analytics.