Iskanje po katalogu digitalne knjižnice

Opis: The present doctoral thesis develops a pioneering system for the self-management of jural eligibilities by means of ICTs as the basis for a novel form of government of juropolitical societies. By means of this system we aim to contribute towards a form of government that would not require a dedicated civil service for the creation, storage, change, and deletion of jural eligibilities in the context of the res publica. In the thesis we explore the concept of jural relations as the atomic links of governmental systems, the composition of the jural relations as such, as well as the role of the jural subjectivity as a crucial component for creating complex systems of jural relations that serve as the underlying structures of juropolitical systems. We then walk through the history of the civil service – the bureaucratic machine, as Banfield called it, to understand its role and implications on the course of civilization, up till present time, where we discuss the impacts of ICTs on the development of the bureaucratic machine as such. We argue that the changes which ICTs so far brought to the government sector through what is known as e-government, e-democracy, and e-governance respectively, are based on unsustainable artefacts and hence there are strong reasons for them to be considered more of a burden to future generations, rather than a source of relief. Based on the implications of the so explored context we describe a model for an information system that would enable self-managed creation and determination of jural eligibilities, and thus self-managed government of juropolitical societies as such. We call this model Sustainable Non-Bureaucratic Government (SNBG). SNBG bases on a network of electronic registries, which store jural facts, from which eligibilities can be derived through a dedicated mechanism, which we call constellation-based reasoning (CBR). CBR bases on a purposely developed fine-grained data access control mechanism, which does not rely on predefined accessor roles, but dynamically enables / disables access to data based on the context of the request and the context of the data stored in the accessed registry. As such, CBR is purposely designed to support changing the rules of access to the stored data by means of collaborative decision making, as such is required in the political legislative context, whereby the rules that regulate such decision making, are again governed by the very same system, which ensures full flexibility of the SNBG system to fluidly undergo at design-time unpredictable transitions that would happen through time. This feature amongst others then, assures the system’s sustainability. We describe the architecture and the stakeholders of SNBG, as well as auxiliary constructs for planning and communicating regulations which make-up the CBR rules. We define the functional characteristics that instances of the electronic registries must satisfy in order to assure sustainability and to be applicable in the juropolitical context in accordance with core jural principles (and in order to avoid the mistakes as conducted in the course of development of e-government artefacts). Then, we describe the instantiation of a prototype SNBG system, i.e. the instantiation of a respective electronic registry that provides CBR-based access to the underlying data stored in a relational database. We evaluate this prototype instantiation based on three demo applications, which prove its technical feasibility in different scenarios. Finally, we evaluate the SNBG model in four different real-world scenarios to argue for its feasibility in crucial governance situations.
Ključne besede: unsustainability of e-government, self-management of jural relations, computability of jural eligibilities, non-bureaucratic government, collaborative decision making, liquid democracy, fine-grained data access control, fair non-repudiation, digital identity
Objavljeno v DKUM: 04.06.2015; Ogledov: 2045; Prenosov: 82
Celotno besedilo (11,06 MB)

Opis: The thesis is comprised of three main categories. The first part of dissertation covers investigations of phase equilibria of compounds from natural materials in conventional and also non conventional supercritical fluids. In details, the impact of pressure and/or temperature on the system behaviour (miscibility, solubility, phase inversion) is investigated, quantitative and qualitative analyses to evaluate and identify compounds contained after performing preliminary extraction experiments from different natural tissues are presented. The impact of operating parameters (pre-treatment of the raw material with SFE; different extraction solvents: propane, CO2, non conventional SCFs; different extraction temperatures and pressures) on extraction kinetics is observed. Following substances were taken into consideration: vanillins, caffeine, carnosoic acid extract and lecithin. Second part of dissertation covers studies of phase equilibria of the systems bio oil/gas, which is crucial in biorefinery process design. In this part of dissertation, which covers studies of phase equilibria of binary and ternary systems, the impact of pressure and/or temperature on the system behaviour (miscibility, solubility, phase inversion) for binary system bio oil/supercritical fluid (bio oil/CO2) and (bio oil/H2) was studied. Additionally, phase behaviour of ternary systems of (bio oil/diesel/CO2) and (bio oil/tail water/CO2) under the impact of pressure and/or temperature is observed. These data are of a high importance for bio refineries as an important part of necessary sustainable development. In recent years, studies on biodiesel synthesis have focused on development of process intensification technologies to resolve some of these issues. Fundamental data to design fractionation process of components of bio oil are crucial for an efficient hydrogenation process of bio oil. In the third part of dissertation observation of phase equilibria and determination of the parameters like diffusion coefficient, density and viscosity for the systems polymer/CO2 at elevated pressures is investigated. An overview of different methods applied to determine the parameters like diffusion coefficient, density and viscosity of the systems polymer (PEG)/CO2 at elevated pressures is offered. Observation of phase equilibria of the binary system PEG/CO2, determination of the impact of pressure and/or temperature on the system behaviour (miscibility, solubility, phase inversion), determination of thermodynamically and physically properties of the system with new applicative methods and finally, comparison of the results obtained by different methods is provided. The interfacial tension (IFT) at the (PEG)/CO2 interface has been determined by using an experimental technique developed to study the interfacial interactions of the liquids in equilibrium with gas in a glass-windowed equilibrium cell by the means of Capillary Rise (CR) method. Advantages and disadvantages of methods that were applied are exposed and discussed.
Ključne besede: phase equilibria, natural materials, conventional and non conventional supercritical fluids, extraction, bio oil, data for biorefinery process design, systems polymer (PEG)/CO2, diffusion coefficient, density, viscosity, surface tension, Capillary Rise (CR) method.
Objavljeno v DKUM: 28.10.2014; Ogledov: 2968; Prenosov: 393
Celotno besedilo (4,46 MB)

Opis: Pristanišča Koper, Rijeka, Trst, Benetke in Ploče imajo pomembno vlogo v mednarodni trgovini, saj je njihov geografski položaj idealen za nadaljnjo distribucijo blaga proti državam Srednje Evrope. Pri tem je delovanje kontejnerskih pristanišč na svetovnem trgu, možno predstaviti z mero učinkovitosti. V magistrskem delu z uporabo metode DEA (ang. Data Envelopment Analysis, sl. analiza ovojnice podatkov) , analizirana je relativna tehnična učinkovitost kontejnerskih pristanišč Koper, Rijeka, Trst, Benetke in Ploče. Uporabljeni so bili različni modeli z različnimi kombinacijami vhodnih in izhodnih spremenljivk. Rezultati kažejo, da na učinkovitost pristanišč najbolj vpliva sorazmerje vhodnih in izhodnih spremenljivk.
Ključne besede: kontejnerska pristanišča, Data Envelopment Analysis (DEA), učinkovitost.
Objavljeno v DKUM: 04.02.2014; Ogledov: 2697; Prenosov: 413
Celotno besedilo (1,83 MB)

Opis: Informacije predstavljajo neotipljivo premoženje organizacij in so njihova pomembna konkurenčna prednost. Za najboljši izkoristek informacij se organizacije posložujejo poslovnega obveščanja, katerega glavna komponenta so podatkovne baze. Najbolj razširjene podatkovne baze so relacijske baze podatkov, pri katerih so podatki shranjeni v logično povezanih tabelah. Trend povečane rasti količine podatkov je povzročil, da relacijske baze ne uspejo zadovoljiti potreb poslovnega obveščanja. Pojavil se je koncept Big Data in z njim povezane tehnologije za zajem, shranjevanje, obdelavo in izkoristek masovnih količin podatkov. Big Data predstavljajo veliko poslovno priložnost tako za organizacije, ki se jih bodo posluževale pri poslovnem obveščanju, kot tudi za ponudnike programske opreme. Orodje Apache Hadoop, ki vsebuje dve pomembnejši komponenti – MapReduce in HDFS, je le ena izmed prepoznavnejših programskih rešitev na področju koncepta Big Data. Pomembna storitev, ki spada k velikim podatkovnim bazam je tudi računalništvo v oblaku, saj zadovoljuje zahteve s strani infrastrukture, prilagodljivosti storitve pa tudi varstva podatkov.
Ključne besede: Big Data, tehnologije za obvladovanje Big Data, Apache Hadoop, MapReduce, HDFS, poslovno obveščanje, podatkovne baze, relacijske podatkovne baze, integracija podatkov, računalništvo v oblaku
Objavljeno v DKUM: 04.11.2013; Ogledov: 2522; Prenosov: 497
Celotno besedilo (974,42 KB)

Opis: Diplomsko delo obravnava podatkovno rudarjenje, metode strojnega učenja, obstoječe aplikacije za podatkovno rudarjenje, ter uporabo in integracijo knjižnice Weka v ogrodje .NET, s katerim smo izdelali aplikacijo za obdelavo podatkov Zavarovalnice Maribor. Na Zavarovalnici Maribor smo dobili odobritev, da lahko pri svoji nalogi uporabimo njihove podatke, ki bodo zaradi varnosti nekoliko okrnjeni. V diplomskem delu smo izdelali in opisali aplikacijo, ki omogoča statistično obdelavo vnesenih podatkov in ki na podlagi podatkovnega rudarjenja ter strojnega učenja omogoča analizo in ekstrakcijo določenih pravil, npr. kakšni zavarovanci so najbolj tvegani za zavarovalnico.
Ključne besede: podatkovno rudarjenje, strojno učenje, odločitvena drevesa, Bayesov klasifikator, nevronske mreže, hibridne metode, ansambelske metode, grobe množice, asociativna pravila, Apriori, ikvm, Data Miner, Mahout, Weka, .NET
Objavljeno v DKUM: 30.11.2012; Ogledov: 3461; Prenosov: 642  (2 glasa)
Celotno besedilo (2,19 MB)