There is currently an effort within the European Union and Sweden to make data produced by government agencies available for reuse to citizens and organizations. While Sweden has had a long history of sharing information with its citizens the country is currently behind its northern neighboring countries when it comes to publishing government data openly. In this study, the author seeks to find if there are types of governmental agencies whose work makes it easier or harder to make it available for reuse. This was done in two phases; in the first phase, the author identified overrepresented and underrepresented agency types on Sweden’s open data portal. In the second phase, the author interviewed two agencies from an overrepresented agency type and two agencies from an underrepresented agency type, to learn what has enabled the overrepresented agencies and what has prevented the underrepresented agencies to publish data openly. The results show that agencies that have sensitive data, in general, had a harder time publishing their data openly than those which did not. The agencies that were underrepresented on the Swedish data portal also had decentralized data generation methods and ways of storage that negatively impacted their ability to publish data openly. What enabled the overrepresented agencies to publish data, beyond having a low amount of sensitive data was the tangible benefits of publishing data openly and the organizational willingness that came with these benefits.

Im Herbst 2012 feiert der Lehrstuhl Datenbanken an der Technischen Universität Dresden sein 10-jähriges Bestehen unter der Leitung von Wolfgang Lehner. In diesem Zeitraum wurde die inhaltliche Ausrichtung im Bereich der Datenbankunterstützung zur Auswertung großer Datenbestände weiter fokussiert sowie auf Systemebene deutlich ausgeweitet. Die Forschungsgruppe um Wolfgang Lehner ist dabei sowohl auf internationaler Ebene durch Publikationen und Kooperationen sichtbar als auch in Forschungsverbünden auf regionaler Ebene aktiv, um sowohl an der extrem jungen und agilen Software-Industrie in Dresden zu partizipieren und, soweit eine Forschungsgruppe dies zu leisten vermag, auch unterstützend zu wirken. [Aus: Einleitung]

The Swedish water sector faces many challenges in terms of climate change, worn out facilities and distribution systems as many of them were built in the 50-, 60- and 70s, but also increased cyber security concerns. Collaborations between different actors are therefore required to address these challenges. This thesis work's purpose is to examine the opportunities a collaboration between water treatment plants could bring for the water sector to achieve a more sustainable drinking water production. The purpose is also to examine the most efficient way to establish a collaboration between them through digital means. A workshop and several interviews with representatives from different water treatment plants were conducted to gather information on possibilities of data/information sharing between them, but also to make a security classification of data/information generated at the water treatment plants. The representatives included process engineers, process technicians, security associates and an associate at Svenskt vatten, an interest organization for the water sector. The result showed that data/information sharing is already taking place in the sector under limited conditions. Smaller VA organizations have more challenges due to lack of skilled workers, partners and time. It is difficult for many in the sector to know where to turn to when in need of help, particularly for smaller VA organizations who usually are the ones with fewer contact networks. It is not to say that only smaller VA organizations are affected but they are more likely to lack contact with other water organizations. The majority of water treatment plants agree on that data/information on treatment processes would be allowed to share under limited conditions. The information classification is a matter for each plant due to different security realities which determines the information classification. This makes it difficult to decide on what parameters could be shared by every water treatment plant. This needs to be investigated further in order to determine what data/information could be shared by most water treatment plants on a platform. Most participants thought that a platform of some kind would be the best solution to simplify data/information sharing between water treatment plants. The suggested platform is a good start for making contacts between water treatment plants and has the potential to enable storage and sharing of data/information in the future. / Den svenska VA sektorn står inför många utmaningar så som klimatförändringar, slitna anläggningar och distributionsnät då många av dem byggdes under 50-, 60- och 70-talet. Även oron kring cybersäkerheten ökar. Samarbeten mellan olika aktörer krävs för att lösa dessa utmaningar. Syftet med examensarbetet är att undersöka de möjligheter som samarbeten mellan vattenverk skulle medföra för att nå en mer hållbar dricksvattenproduktion. Syftet är också att undersöka det mest effektiva sättet att få till ett samarbete mellan vattenverken med digitala medel. En workshop och flertalet intervjuer genomfördes med olika representanter från vattenverk för att samla information om möjligheten med att dela data/information mellan dem. Men också för att göra en säkerhetsklassning av data/information som genereras på vattenverken. Representanterna inkluderade processingenjörer, processtekniker, säkerhetsansvariga samt en medarbetare från Svenskt vatten, en intresseorganisation för vattensektorn. Resultatet visade att delning av data/information redan sker inom sektorn, dock under begränsade former. Mindre VA organisationer har fler utmaningar på grund av brist på anställda med rätt kompetens, samarbetspartners och tid. Många inom sektorn har utmaningar när det kommer till att söka och få hjälp, särskilt för de mindre VA organisationerna som vanligtvis har färre kontakter. Men de är inte bara de små som är påverkade av detta problem, men det är vanligare att de inte har kontakter med andra VA organisationer. De flesta vattenverken ansåg att data/information från reningssteg skulle vara möjligt att dela under begränsade former. Informationssäkerhetsklassning är en enskild fråga för varje vattenverk som baseras på de hot som varje vattenverk står inför. Detta medför att det är svårt att bestämma vilken data/information som kan delas av varje vattenverk. Detta område behöver vidare utredas innan ett beslut kan fattas om vilken data/information alla vattenverk ska dela på en plattform. De flesta deltagarna tyckte att en plattform av något slag skulle vara lösningen för att förenkla delning av data/information mellan vattenverken. Den föreslagna plattformen är en bra start för vattenverk att skapa kontakter och har potentialen att möjliggöra lagring och delning av data/information i framtiden.

No description available.

No description available.

Platforms for publication and collaborative management of data, such as Data.gov or Google Fusion Tables, are a new trend on the web. They manage very large corpora of datasets, but often lack an integrated schema, ontology, or even just common publication standards. This results in inconsistent names for attributes of the same meaning, which constrains the discovery of relationships between datasets as well as their reusability. Existing data integration techniques focus on reuse-time, i.e., they are applied when a user wants to combine a specific set of datasets or integrate them with an existing database. In contrast, this paper investigates a novel method of data integration at publish-time, where the publisher is provided with suggestions on how to integrate the new dataset with the corpus as a whole, without resorting to a manually created mediated schema or ontology for the platform. We propose data-driven algorithms that propose alternative attribute names for a newly published dataset based on attribute- and instance statistics maintained on the corpus. We evaluate the proposed algorithms using real-world corpora based on the Open Data Platform opendata.socrata.com and relational data extracted from Wikipedia. We report on the system's response time, and on the results of an extensive crowdsourcing-based evaluation of the quality of the generated attribute names alternatives.

Den första detaljerade topografiska karteringen som täcker en större del av Sverige är den s.k. Häradskartan som framställdes under 1800-talets andra hälft och som följdes av Generalstabens topografiska karta under början av 1900-talet. För framställning av kartorna upprättades först stomkartor genom transporter av äldre kartor i större skala från skiften och lantmäteriförrättningar. Historiska topografiska kartor ur Häradskartan och stomkartorna till Generalstabens topografiska karta publicerades under 2023 av Naturvårdsverket i form av digitala kartraster som georefererats med en automatisk metod baserad på rutnät av kartbladshörn för kartserierna. Denna studie syftar till att undersöka huruvida resultatet av den automatiska georefering som utförts med kartbladshörnen kan förbättras med en metod baserad på identifiering av gemensamma passpunkter mellan de historiska kartan och den moderna topografiska karta som finns tillgänglig som öppen data från Lantmäteriet. För att den metod som används ska vara användbar för en intresserad allmänhet undersöktes de möjligheter för georeferering av de historiska topografiska kartorna som finns med kostnadsfria Open Source-programvaror och öppna data. För att resultatet av georefereringen ska kunna utvärderas i en praktisk tillämpning undersöktes skapande av 3D-visualiseringar av de historiska kartorna. Studien avgränsas till ett område omfattande 400 000 ha huvudsakligen i Örebro län och de georefereringsmetoder som finns i programvaran QGIS. Passpunkter identifierades i den moderna topografiska kartan där byggnadspunkter för slott, kyrkor och herrgårdar som kunde verifieras genom sökningar i bebyggelseregister kompletterades med gränspunkter, vägkorsningar och hydrografiska objekt som visuellt kan antas sammanfalla vid studie av geometrier i de automatiskt georefererade kartorna. För att de kartblad som bygger upp kartserierna ska passa ihop över skarvarna efter georeferering mosaikerades dessa kartblad till ett gemensamt virtuellt kartraster. Resultaten av georefereringen visade att de medelavvikelser som uppmättes i de automatiskt rektifierade kartorna, ca 50 m för Häradskartan och ca 150 m för stomkartorna, kunde förbättras till ca 30 m för båda kartmaterialen med ca sex passpunkter/kvadratmil. Det framgår dock att stora lokala variationer i storleken på felen finns såväl före som efter transformationerna och studien visar på att en mer komplex lokal transformationsmetod med en stor mängd passpunkter fördelade över kartan behövs för att erhålla generella förbättringar av passningen. De skapade 3D-visualiseringarna visade på att en visuellt bättre passning ändå kunde erhållas även med de relativt få passpunkter som använts här. / The first detailed topographic mapping of larger parts of Sweden is the so-called District Economic Map that was produced during the second half of the 19th century and which was followed by the General Staff Map in the beginning of the 20th century. These maps were produced from skeleton maps that were constructed from transports of older large-scaled maps from legal shifts and parcel acts. Historical topographic maps from the District Economic Map and the skeleton maps for the General Staff Map were published in 2023 as digital raster maps that has been georeferenced using an automatic method based on a grid of corner coordinates for the map series. This study aims to investigate whether the result of the automatic georeferencing method can be improved using a method of identifying common ground control points between the historical maps and the modern topographic maps that are available as open data from the Swedish Land Survey. In order for the applied methods to be available to an interested general public this study investigates the possibilities of performing the georeferencing using free open-source software and open data. To allow the evaluation of the georeferencing results in a practical application the historical maps are visualized in 3D. The study is limited to an area encompassing 400 000 ha mainly in Örebro county and the georeferencing algorithms that are available in the QGIS software. Ground control points were identified in the modern topographic map by supplementing building points representing castles, churches and manor houses that can be verified by researching registries of built heritage with boundary points, road crossings and hydrographic objects that visually appear to coincide when studying the geometries of the automatically georeferenced maps. In order to ensure that the map sheets that make up the map series fit over the sheet edges the sheets were mosaicked in a single virtual raster. The results of the georeferencing process show that the mean deviations that could be measured in the automatically referenced maps (approx. 50 m for the District Economic Map and approx. 150 m for the skeleton maps) could be improved to approx. 30 m in both cases using about six control points per 10 000 ha. Large variations in the sizes of local errors are prevalent both before and after the transformations and the study shows that a more complex local transformation method using a large amount of control points is required in order to achieve a general improvement of the fit with modern data. Visualizing in 3D did show, however, that a visually significant improvement could be achieved already using relatively few control points as in this study.

This paper describes problems of eGovernment and proposes way to measure e-Government services. Described is the state of the Czech eGovernment in detail to selected projects and in international comparison. The description shows that the development is still ongoing, but in a much slower pace than in the electronically advanced countries. Main identified problem is inconsequent strategic management and resortism, which is constraining deployment of cross-resort projects. Based on the description of the current situation and on the international experiences, author offers specific recommendations to future development of the electronization of public administration in the Czech Republic. In response on management problems is in this paper proposed and tested methodology for measuring the effectiveness of e-Government services. This methodology is an objective procedure of measuring the effectiveness of e-services which is providing feedback for service providers. Testing of methodology evaluates three services from taxation agenda and highlights their weaknesses, including the proposal of corrective solutions.

This thesis presents the implementation of data fusion techniques for sustainable urban development. Recently, increasingly more geospatial data have been made easily available for no cost. The immeasurable quantities of geospatial data are mainly from four kinds of sources: remote sensing satellites, geographic information systems (GIS) data, citizen science, and sensor web. Among them, satellite images have been mostly used, due to the frequent and repetitive coverage, as well as the data acquisition over a long time period. However, the rather coarse spatial resolution of e.g. 30 m for Landsat 8 multispectral images impairs the application of satellite images in urban areas. Even though image fusion techniques have been used to improve the spatial resolution, the existing image fusion methods are neither suitable for sharpening one band thermal images nor for hyperspectral images with hundreds of bands. Therefore, simplified Ehlers fusion was developed. It adds the spatial information of a high-resolution image into a low-resolution image in the frequency domain through fast Fourier transform (FFT) and filter techniques. The developed algorithm successfully improved the spatial resolution of both one band thermal images as well as hyperspectral images. It can enhance various images, regardless of the number of bands and the spectral coverage, providing more precise measurement and richer information. To investigate the performance of simplified Ehlers fusion in practical use, it was applied for urban heat island (UHI) analysis. This was done by sharpening daytime and nighttime thermal images from Landsat 8, Landsat 7, and the Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER). The developed algorithm effectively improved the spatial details of the original images so that the temperature differences between agricultural, forest, industrial, transportation, and residential areas could be distinguished from each other. Based on that, it was found that in the study city the causes of UHI are mainly anthropogenic heat from industrial areas as well as high temperatures from the road surface and dense urban fabric. Based on this analysis, corresponding mitigation strategies were tailored. Remote sensing images are useful yet not sufficient to retrieve land use related information, despite high spatial resolution. For sustainable urban development research, remote sensing images need to be incorporated with data from other sources. Accordingly, image fusion needs to be extended to broader data fusion. Extraction of urban vacant land was therefore taken as a second application case. Much effort was spent on the definition of vacant land as unclear definitions lead to ineffective data fusion and incorrect site extraction results. Through an intensive study of the current research and the available open data sources, a vacant land typology is proposed. It includes four categories: transportation-associated land, natural sites, unattended areas or remnant parcels, and brownfields. Based on this typology, a two-level data fusion framework was developed. On the feature level, sites are identified. For each type of vacant land, an individual site extraction rule and data fusion procedure is implemented. The overall data fusion involves satellite images, GIS data, citizen science, and social media data. In the end, four types of vacant land features were extracted from the study area. On the decision level, these extracted sites could be conserved or further developed to support sustainable urban development.

[ES] La explotación de datos de salud ha demostrado ser de creciente interés en la comunidad científica, especialmente para la creación y uso de sistemas de ayuda a la decisión clínica (SADC). Para abordar este problema, tradicionalmente se ha investigado por separado en materia de modelos de información, modelos de dominio y SADC. En lo que se refiere a modelos de información, las propuestas presentan limitaciones semánticas y no tienen en cuenta la interacción con los modelos de dominio, que pretenden proporcionar una comprensión formal y compartida del conocimiento clínico, ni con los SADC, cuya finalidad es proporcionar apoyo a la toma de decisión clínica a partir de la historia clínica electrónica (HCE). La finalidad de esta tesis se enmarca dentro del objetivo general de enriquecer sistemas de HCE con resultados de SADC y modelos de dominio representados mediante datos enlazados abiertos. Para ello, se investiga la combinación y explotación conjunta de las tecnologías más avanzadas para modelos de información, modelos de dominio y SADC. La principal contribución de esta tesis es el desarrollo de metodologías y herramientas para enriquecer la HCE con resultados de SADC y datos enlazados abiertos. Las contribuciones específicas son las siguientes: * Definición conceptual y metodológica de la HCE aumentada con información potencialmente relevante de la web semántica. * Definición conceptual y metodológica del informe radiológico estructurado (IRE), enriquecido con resultados de SADC basados en reglas, visión por computación y modelos de aprendizaje automático. * Caso de uso de HCE aumentada, consistente en enriquecer la HCE resumida del Sistema Nacional de Salud de España con datos enlazados abiertos sobre interacciones farmacológicas y tratamientos recomendados para los episodios activos del paciente. * Sistema de IRE enriquecido con resultados de SADC. Incluye el desarrollo de plantillas de IRE y mecanismos para el enriquecimiento de las mismas con resultados de SADC basados en reglas, cuantificación de imagen médica y redes neuronales. Nuestro objetivo es mejorar el grado de interoperabilidad en las integraciones de sistemas de HCE con SADC y datos enlazados abiertos, mediante estrategias basadas en los tres pilares de la interoperabilidad semántica: modelos de información, de arquetipos y de dominio. Esto tiene el potencial de repercutir positivamente sobre la salud y el cuidado del paciente, especialmente en el paradigma de la medicina personalizada. / [CA] L'explotació de dades de salut ha demostrat ser de creixent interés en la comunitat científica, especialment per a la creació i ús de sistemes d'ajuda a la decisió clínica (SADC). Per a abordar aquest problema, tradicionalment s'ha investigat per separat en matèria de models d'informació, models de domini i SADC. Pel que fa a models d'informació, les propostes presenten limitacions semàntiques i no tenen en compte la interacció amb els models de domini, que pretenen proporcionar una comprensió formal i compartida del coneixement clínic, ni amb els SADC, la finalitat dels quals és proporcionar suport a la presa de decisió clínica partint de la història clínica electrònica (HCE). La finalitat d'aquesta tesi s'emmarca dins de l'objectiu general d'enriquir sistemes de HCE amb resultats de SADC i models de domini representats mitjançant dades enllaçades obertes. Per a això, s'investiga la combinació i explotació conjunta de les tecnologies més avançades per a models d'informació, models de domini i SADC. La principal contribució d'aquesta tesi és el desenvolupament de metodologies i eines per a enriquir la HCE amb resultats de SADC i dades enllaçades obertes. Les contribucions específiques són les següents: * Definició conceptual i metodològica de la HCE augmentada amb informació potencialment rellevant de la web semàntica. * Definició conceptual i metodològica de l'informe radiològic estructurat (IRE), enriquit amb resultats de SADC basats en regles, visió per computació i models d'aprenentatge automàtic. * Cas d'ús de HCE augmentada, consistent a enriquir la HCE resumida del Sistema Nacional de Salut d'Espanya amb dades enllaçades obertes sobre interaccions farmacològiques i tractaments recomanats per als episodis actius del pacient. * Sistema de IRE enriquit amb resultats de SADC. Inclou el desenvolupament de plantilles de IRE i mecanismes per a l'enriquiment de les mateixes amb resultats de SADC basats en regles, quantificació d'imatge mèdica i xarxes neuronals. El nostre objectiu és millorar el grau d'interoperabilitat en les integracions de sistemes de HCE amb SADC i dades enllaçades obertes, mitjançant estratègies basades en els tres pilars de la interoperabilitat semàntica: models d'informació, d'arquetips i de domini. Això té el potencial de repercutir positivament sobre la salut i la cura del pacient, especialment en el paradigma de la medicina personalitzada. / [EN] The exploitation of health data has proven to be of increasing interest in the scientific community, especially for the creation and use of clinical decision support systems (CDSS). To address this problem, separate research has traditionally been done on information models, domain models and CDSS. Regarding information models, the proposals present semantic limitations and do not consider the interaction with domain models, which aim to provide a formal and shared understanding of clinical knowledge, nor with CDSS, whose purpose is to provide clinical decision support from the electronic health record (EHR). The aim of this thesis is framed within the general goal of enriching EHR systems with SADC results and domain models represented by open linked data. For this purpose, the combination and joint exploitation of state-of-the-art technologies for information models, domain models and SADC is investigated. The main contribution of this thesis is the development of methodologies and tools to enrich EHR with SADC results and open linked data. Specific contributions are: * Conceptual and methodological definition of EHR augmented with potentially relevant information from the semantic web. * Conceptual and methodological definition of the structured radiology report (SRR), enriched with CDSS results based on logical rules, computer vision and machine learning models. * Augmented EHR use case, consisting of enriching the summarized EHR of the Spanish National Health System with linked open data on pharmacological interactions and recommended treatments for active patient episodes. * SRR system enriched with CDSS results. Includes the development of SRR templates and mechanisms for enriching them with SADC results based on logical rules, medical image quantification and neural networks. Our goal is to improve the degree of interoperability in EHR system integrations with CDSS results and linked open data, through strategies based on the three pillars of semantic interoperability: information, archetype and domain models. This has the potential to positively impact health and patient care, especially in the personalized medicine paradigm. / Mañas García, A. (2022). Enriquecimiento de la historia clínica electrónica con información de sistemas de ayuda a la decisión clínica y datos enlazados abiertos [Tesis doctoral]. Universitat Politècnica de València. https://doi.org/10.4995/Thesis/10251/187730 / TESIS