A Distributed and Parallel Processing Framework for Knowledge Graph OLAP

Business intelligence and analytics refers to the ensemble of tools and techniques that allow organizations to obtain insights from big data for better decision making. Knowledge graphs are increasingly being established as a central data hub and prime source for BI and analytics. In the context of BI and analytics, KGs may be used for various analytical tasks; the integration of data and metadata in a KG potentially facilitates interpretation of analysis results. Knowledge Graph OLAP (KG-OLAP) adapts the concept of online analytical processing (OLAP) from multidimensional data analysis for the processing of KGs for analytical purposes. The current KG-OLAP implementation is a monolithic system, which greatly inhibits scalability. We propose a research plan for the development of a framework for distributed and parallel data processing for KG-OLAP over big data. In particular, we propose a framework for KG-OLAP over big data based on the data lakehouse architecture, which leverages existing frameworks for parallel and distributed data processing. We are currently at an early stage of our research.