Thinking the Incorporation of LOD in Semantic Cubes as a Strategic Decision

Abstract

With the advent of Linked Open Data (LOD) initiatives, organizations have seen the opportunity of augmenting their internal data cube systems with these external data. While IT actors manage technical issues of internal and external sources, a new actor has emerged “the Chief Data Officer” (CDO), the role of which is to align and prioritize data activities with key organizational priorities and goals. Existing literature managing the incorporation of LOD in internal Data cubes mainly focus on technical aspects of the LOD source and ignores the CDO role in this strategy. In this paper, we claim that technical actions should be conducted by the managerial level, which is reflected through the goals of the organization data cube and their related Key Performance Indicators (KPIs). For doing this, we first propose a metamodel aligning the three models: the data-flow model, the goal model and the KPI model. Then, we propose a process for specifying KPIs into Sparql language, the standard language for querying LOD sources. Experiments are conducted to measure the impact of the decision of integrating external LOD sources at KPI/goal level and on the technical data level. A case tool dedicated to the CDO is implemented to conduct the proposed approach.