A user-centric framework to improve the reusability
A thesis submitted in partial fulfillment of the requirements for the degree of Doctor in Information Management, specialization in Geographic Information Systems ; Open data has a profound effect in working environments in which information is created and shared at all levels. At the local government level, open-data initiatives have resulted in higher levels of transparency as regards policies. Greater engagement between decision-makers and citizens has changed the way data analysis and evidence are used to support local governance. Initiatives on open data are currently playing an essential role in local governments. However, the current challenge of local open data that authorities are facing has gradually changed from accessibility issues to measures of the impact of the ongoing open-data projects, from more data catalogs to sustainable and increasing levels of reuse of released data, and better reusability of open data. Despite an increasing amount of data being made open, few studies have looked into its level of reusability, and the barriers that hamper the reuse of open geodata from a data consumer's perspective are an issue that most communities of data users are currently faced with. Some frameworks are showing how the level of maturity in national open-data initiatives is either increasing or decreasing, but there is still a need for a specific framework to guide local data authorities to engage their current users and also help them to move toward a bottom-up approach. This research contributes with three elements in this regard. The first is the current status of the level of reuse of open geodata in cities. This is followed by a taxonomy of the barriers faced by data users in Colombia and Spain, and the third is a set of elements that shape a user-centric framework to help data authorities improve the level of reuse of published open geodata in their ongoing local initiatives. The proposed taxonomy and framework are based on a literature review, an online survey, and a set of participatory workshops conducted in four selected cities (Bogotá, Medellín, Cali in Colombia and Valencia in Spain), with local data authorities and user communities from different backgrounds and with experience in the field of open data. The taxonomy presented in this research highlights a number of issues such as outdated data, low integration of data producers, and difficulty to access data, the most relevant from the data consumer's point of view being misinterpretation and misuse of released data and their terms of use. Once the barriers had been identified and validated with data users across the selected cities, this research defined the elements included in a conceptual framework that local authorities could use as a guideline to improve the level of reuse in their ongoing open data initiatives. The core elements of this framework are what are defined as 'Impact Enablers', which consist of three aspects considered by the literature reviewed as relevant to improve the positive impact of current initiatives. These three factors are: A) the requirements of data-user communities; B) open data at city level as a way to promote and engage users; and finally, C) a geographic approach to improving the level of reusability of released data due to its potential to engage more users. The second part of the proposed framework is made up of four connected elements: 1) The complete identification of data-user communities and their needs; 2) The community of reuse as a set of technological tools to promote the reusability of released data; 3) User-focused metadata; and 4) Reuse-focused legal terms. The elements mentioned earlier were compiled and included due to their relevance for data-user communities in the four use cases included in this research. This framework provides a clear path for local data authorities to reshape their current open data strategies so as to include data-user requirements and move toward a bottom-up approach. The research ends with a discussion and some concluding points, in addition to several limitations in the application of our findings. At the end of this dissertation, a roadmap for future research and implementations are presented, taking into account some reflections on the framework.