The whole work is dealing with so-called. OpenData. Since this topic is very extensive, this paper focuses only on cataloging tools (catalogs) for open data. The main reason for this narrower focus is still great variability and new technologies that are of interest not only for the individuals or institutions but also for the whole public administration. For example at the beginning of this work none Czech national open data catalog existed, in the course of writing it managed to launch and now works almost flawlessly. Therefore, it is an important part of a retrospective view. The main objective of this work is to analyze the situation of the national catalogs of open data in selected locations, mapping the requirements of catalogs by avaible documents, creation of criteria and assessed according to criteria selected cataloging tool. Benefits include compiled criteria that can serve additional work or other purposes, and the aforementioned retrospective view. The structure of the work will be divided into several basic blocks, which can be summarized into the structure: Introduction, practical part, analytical part, conclusion. The introduction explains the reasons and aims of work and achievement. The practical part describes the past and present status issues. In the analytical part are assessed as suitable criteria / requirements that are applied to the selected tool. In the end everything is summed up.

Das Ziel der Studie ist die Entwicklung eines Fehlerkorrekturmodells fürWettervorhersagedaten, welches zu einem besseren Verständnis der Zuverlässigkeit der lokalenWettervorhersagen beitragen soll. Weiterhin soll das Fehlermodell in weiterführenden Arbeiten innerhalb eines stochastischen modellprädiktiven Reglers (MPC) Anwendung finden. Im Rahmen der Studie werden sowohl prognostizierte als auch tatsächliche Wetterdaten aus dem Raum Fulda erfasst. Diese Daten werden grafisch dargestellt und mittels Fourier-Transformation analysiert. Auf Basis dieser Analyse werden mathematische Modelle zur Vorhersage von Fehlern und zur Bestimmung der Varianz vorgeschlagen. Die Parameter dieser Modelle werden durch lineare Regression bestimmt. Die erzielten Ergebnisse werden ausführlich ausgewertet.

Depuis les années 2000, les technologies du Web permettent aux utilisateurs de prendre part à la production de données : les internautes du Web 2.0 sont les nouveaux capteurs de l’information. Du côté de l’Information Géographique affluent de nombreux jeux de données en provenance de plates-formes de cartographie participative telles qu’OpenStreetMap (OSM) qui a largement impulsé le phénomène de la Géographique Participative (VGI). La communauté OSM représente aujourd’hui plus de deux millions de contributeurs qui alimentent une base de données géospatiales ouverte dont l’objet est de capturer une représentation du territoire mondial. Les éléments cartographiques qui découlent de ce déluge de VGI sont caractérisés par des tags. Les tags permettent une catégorisation simple et rapide du contenu des plates-formes de crowdsourcing qui inondent la toile. Cette approche est cependant un obstacle majeur pour le partage et la réutilisation de ces grands volumes d’information. En effet, ces ensembles de tags, ou folksonomies, sont des modèles de données beaucoup moins expressifs que les ontologies. Nous proposons un méta-modèle pour rapprocher la folksonomie et l’ontologie OSM afin de mieux exploiter la sémantique des données qui en sont issues, tout en préservant la flexibilité intrinsèque à l’utilisation de tags. / Post-2000s web technologies have enabled users to engage in the information production process: Web 2.0 surfers are the new data sensors. Regarding Geographic Information (GI), large crowdsourced datasets emerge from the Volunteered Geographic Information (VGI) phenomenon through platforms such as OpenStreetMap (OSM). The latter involves more than two millions contributors who aim at mapping the world into an open geospatial database. This deluge of VGI consists of spatial features associated with tags describing their attributes which is typical of crowdsourced content categorization. However, this approach is also a major impediment to interoperability with other systems that could benefit from this huge amount of bottom-up data. Indeed, folksonomies are much less expressive data models than ontologies. We address the issue of loose OSM metadata by proposing a model for collaborative ontology engineering in order to semantically lift the data while preserving the flexible nature of the activity of tagging.

Transport models are key in predicting travel behaviour and planning transport systems. Transport models can be either aggregated or disaggregated. Disaggregation means that travel behaviour is represented on an individual level, which can be beneficial because it offers a higher detail level and reduces aggregation bias. Input data for transport models can be both expensive and inaccessible, especially comprehensive data. Thus, it is advantageous to explore the utilisation of open data, which is free and accessible. The objective of the thesis was to evaluate how OpenStreetMap and other Open Data can be utilised in disaggregated transport modelling. The scope of the study was Uppsala, Sweden. In the thesis, a disaggregate transport model was designed, which only considered commuting trips made by public transport. Destinations and a synthetic population were estimated based on OpenStreetMap map features, SCB census data, and LuTRANS land use data. A travel survey was utilised in model calibration, and UL boarding data was used for model validation. The results showed that OpenStreetMap provided sufficient data for estimating a synthetic population and destinations for a disaggregate transport model when combined with other open data sources. Population and land usecensus data were essential for calibrating the model. However, the model came with limitations caused by assumptions, generalisation, technical constraints, and the partial incompleteness of open data. The thesis concludes that Open Data, such as OpenStreetMap, can be utilised sufficiently for transport modelling, with proper assumptions and processing. The openness of the data also increases the replicability of such a model. / Transportmodeller är viktiga i att förutspå resvanemönster och för att kunna planera transportsystemet. Transportmodeller kan vara antingen aggregeradeeller disaggregerade. Disaggregering betyder att resvanor är representerade påindividuell nivå, vilket kan vara fördelaktigt då det innebär en högre detalj nivå och mindre partiskhet orsakad av aggregering (aggregation bias). Indata förtransportmodeller kan vara både dyrt och svåråtkomligt, speciellt för mer omfattande data. Därav kan det vara till stor nytta att utforska möjligheten att använda öppnadata (Open Data), som är gratis och lättåtkomligt. Syftet med examensarbetetvar att utvärdera hur OpenStreetMap och annan Open Data kan användas idisaggregerad transportmodellering. Den geografiska omfattningen av studien är Uppsala tätort. En disaggregerad transportmodell togs fram i examensarbetet, sombara tog hänsyn till jobbresor med kollektivtrafik. Destinationer och en syntetiskbefolkning uppskattades utifrån OpenStreetMap objekt, befolkningsdata från SCB, samt markanvändningsdata från LuTRANS. En resvaneundersökning utnyttjadesför modellkalibrering och påstigningsdata från UL användes för modellvalidering.Resultaten visade att OpenStreetMap erbjöd tillräckligt med data för att ta framoch uppskatta en syntetisk befolkning och destinationer för en disaggregeradtransportmodell, om den kombineras med andra öppna datakällor. Befolkning- ochmarkanvändningsdata var avgörande i att kalibrera modellen. Dock så innefattar modellen vissa begränsningar som är orsakada av antaganden, generalisering, tekniskabegränsningar, samt ofullständigheten av Open Data. Slutsatsen är att Open Data, så som OpenStreetMap, kan utnyttjas för transportmodellering, om det kombineras med välformulerade antaganden och processering av datan. Datans öppenheten medför även en ökad replikerbarhet för en sådan modell.