Παπαγεωργίου Γεώργιος

Περίληψη:
This study evaluates the efficacy of Large Language Models (LLMs) for the enrichment of migration data using unstructured, fragmented open government data and journalistic articles. By benchmarking six models against a manually compiled journalistic dataset of 129 incidents of migration on the island of Crete during the first two quarters of 2025, we examine the extraction of information such as refugee demographics, perpetrator profiles, and smuggling fees. The results reveal that the models were highly capable of demographic extraction, but their effectiveness diminished when faced with more complex tasks that required geographic reasoning and the synthesis of financial information. The findings of the study conclude that the use of LLMs can be a viable alternative to bespoke software for investigative reporting, provided there is thorough human oversight.

Περίληψη:
The development of ecosystems around open government data in order to generate social and economic value from them requires on one hand the existence of a critical mass of them and on the other hand some level of homogeneity and interoperability among the numerous portals and websites providing them which enables their combined exploitation. This paper presents the Technical, Semantic, Legal, and Organizational (TSLO) Interoperability Framework, a conceptual tool designed to provide a structured evaluation of the interoperability capabilities of open government data infrastructures. It incorporates the technical, semantic, legal, and organizational aspects of open government data provision, which are crucial for developing interoperability between open data portals and websites, which enables the generation of high levels of social and economic value from them, so it can pave the way for creating open data ecosystems for this purpose. We are then applying this TSLO framework to examine from this perspective the open data infrastructures of the Greek local government administration institutions of all three layers of it (decentralized, regional, and municipal). In particular, we examine the open government data portals and websites of the local Greek governments and provide a detailed analysis of them concerning the fore aspects of the framework. This first application of TSLO reveals not only the low participation of Greek local government institutions in the open data movement, and especially of the municipalities, where only 8 out of 332 municipalities provide open data on their portals or websites but also the inconsistent implementation of this, especially with respect to the use of semantic technologies as well as the legal frameworks under which the open data provided. Finally, we propose the automation of the TSLO framework to facilitate the evaluation and standardization of open local government data infrastructure and expand the study to identify and address barriers to open data provision.

Περίληψη:
Whilst there is a need for furthering the engagement of the universities in imbibing the Open Government Data (OGD) curriculum in their graduate and postgraduate courses, there has been no systematic research focused on the identification of the qualitative benchmarking regarding the curriculum of OGD-focused academic programmes. Benchmarking the OGD curriculum with the adoption of best practices among the universities is pertinent given the significant role of the students and other faculty members in the value derivation and innovation pursuits in their current and futuristic capacities. Drawing inferences from the QS World University rankings, 2023, via the BEKA (benchmarking, evidencing, knowing, applying) analysis, this paper identifies the best practices in qualitative terms for furthering dialogue and discourse among the university stakeholders and others in the OGD ecosystem. Specifically, a detailed course structure is provided for OGD courses which may serve as benchmarks for further foundation-laying of OGD curriculum in different varsities. As a first contribution towards identification of the best practices in the OGD-focused curriculum designing, the study contributes towards OGD literature, in general, and the benchmarking practices in OGD-focused discourse, in specific

Περίληψη:
Extant research on Open Government Data (OGD) has remained confined to the grappling of issues linked with its conceptual, theoretical and empirical dimensions, however focusing on the supply of OGD physical capital (ODG portals, datasets, etc.), and to a lesser degree on the demand for it (e.g., needs of potential users), but not dealing with relevant OGD human capital (human knowledge and skills concerning OGD). Furthermore, research on meta-analysis or literature reviews has not expanded its scope to unravel the formation of OGD human capital, and especially how the OGD theme is being showcased across universities’ curriculum. The present research aims to contribute to filling this research gap, through an analysis of the OGD-related programs and courses offered at the graduate and post-graduate levels across the top-notch universities identified as per the indicators of the QS World University rankings, 2023. Our theoretical foundation is the widely recognized ‘Human Capital Theory’ from the economic science, which gives prominence to the importance of the human capital (human knowledge and skills) as an important complement of the classical physical capital (e.g., production equipment, ICT capital, etc.). Our findings indicate that there are only small number of courses concerning OGD in these top-notch universities; furthermore, a very small share of them have OGD an main topic, while most of them include only a part concerning OGD. Most of them focus on the exploitation of OGD of a specific thematic domain (mainly urban studies and health), and only a few deal with OGD in general. Furthermore, there is a prevalence of postgraduate courses, offered as part of MSC programs, followed by undergraduate courses, offered as part of BSC study programs, and to a much lesser degree short courses. Also, with respect to the objective of these OGD-related programs and courses, most of them aim at the generation of scientific value from OGD, while a smaller number aim at the generation of social-political value, and only a much smaller number at the development of economic value. Therefore, it can be concluded that the formation of OGD human capital by the examined universities is limited.