An Exploratory Empirical Assessment of Italian Open Government Data Quality With an eye to enabling linked open data

Context The diffusion of Linked Data and Open Data in recent years kept a very fast pace. However evidence from practitioners shows that disclosing data without proper quality control may jeopardize datasets reuse in terms of apps, linking, and other transformations. Objective Our goals are to understand practical problems experienced by open data users in using and integrating them and build a set of concrete metrics to assess the quality of disclosed data and better support the transition towards linked open data. Method We focus on Open Government Data (OGD), collecting problems experienced by developers and mapping them to a data quality model available in literature. Then we derived a set of metrics and applied them to evaluate a few samples of Italian OGD. Result We present empirical evidence concerning the common quality problems experienced by open data users when using and integrating datasets. The measurements effort showed a few acquired good practices and common weaknesses, and a set of discriminant factors among datasets. Conclusion The study represents the first empirical attempt to evaluate the quality of open datasets at an operational level. Our long-term goal is to support the transition towards Linked Open Government Data (LOGD) with a quality improvement process in the wake of the current practices in Software Quality.