Towards defining data interpretability in open data portals: Challenges and research opportunities

Abstract

Open data portals are growing in scope, and the development of this initiative remains one of the main ways to help create new value for society and the economy. Citizens can use the open data made available on these portals to participate more effectively in democratic processes. For that, they have to be able to access, manipulate and interpret such data. Different authors present different definitions and perceptions about the meaning of data interpretability. Today there is no formal consensus on the concept of data interpretability. The goal of this work is to conceptualize what data interpretability is formal. For this, we carried out literature research to identify the definitions of data interpretability. In addition, we studied the information quality literature to identify the Non-Function Requirements that shape the concept of information quality. So, we aligned the interpretability characteristics with the NFR Framework characteristics to find a unique definition. We also conduct a qualitative analysis with experts in data analysis, e-government, and transparency to identify what these experts understand by interpretability. Based on these two studies, we defined interpretability through a model composed of 8 dimensions, each consisting of different characteristics, which must be guaranteed in the data interpretability process to interpret the data correctly. We understand that, for such characteristics to being guaranteed in the interpretability of open government data, it is necessary to have computational tools to support the user. Thus, we also surveyed which technologies and methods ensure each of the interpretability characteristics and pointed out which computational tools implement such technologies and methods. Finally, we analyzed three large open data portals to identify which characteristics are present in these portals, and we note that there are still several challenges to be handled in open government data portals.