20.500.12556/DKUM-63111 ANALIZA IN NADGRADNJA INFORMACIJSKEGA SISTEMA O RAZISKOVALNI DEJAVNOSTI S SEMANTIČNO KOMPONENTO Analysis of current research information systems and proposal of updating them with a semantic component V disertaciji je predstavljena problematika dostopa do relevantnih informacij v heterogenem okolju informacijskih sistemov o raziskovalni dejavnosti (CRIS sistemov). Na nekaterih karakterističnih primerih, ki so bili povzeti iz prakse slovenskega CRIS sistema SICRIS je uvodoma pokazano, da je zaradi slabega dostopa do relevantnih informacij smiselna uvedba ontologij predvsem v sistemih, ki črpajo informacije iz heterogenih relacijskih baz. Izboljšanje dostopa do relevantnih informacij je predlagano s pomočjo nadgradnje informacijskega sistema s semantično komponento. V ta namen je bil zamišljen poseben model ontološke infrastrukture, za katero je bila napisana namenska testna ontologija z orodjem Protege. V nalogi se prav tako posebej izpostavlja nov, poenostavljen algoritem kreiranja ontologije (vpisa ontoloških razredov), ki ima za osnovo ločitev med vsebinsko novimi in že obstoječimi razredi. Ontologija, zapisana po tem algoritmu je namenska in služi predvsem za potrebe iskalnika. Njena glavna prednost je v tem, da je zaradi opisanega postopka kreiranja ontologija manj obsežna, s čemer se občutno poenostavi postopek njenega kreiranja. Model ontološke infrastrukture in tudi algoritem kreiranja ontologije imata namreč za cilj čim učinkovitejšo podporo procesu iskanja relevantnih rezultatov. Podpora za povečevanje relevantnosti iskalnih rezultatov se pri predlaganem modelu izvede s fasetiranim načinom zoževanja obstoječega iskalnega rezultata, pri tem pa zoževanju služijo novi metapodatki, ki se generirajo iz ontologije. S tako dobljenimi metapodatki se lahko upravičeno pričakuje večja učinkovitost uporabnikov pri iskanju relevantnih informacij, saj se pri opisani metodi v okviru obstoječega priklica poveča preciznost iskanja. Za navedeni model je bila na koncu v orodju Protege izvedena še simulacija procesa iskanja s pomočjo poenostavljene ontologije slovenske raziskovalne dejavnosti. This dissertation presents the issue of access to relevant information in a heterogeneous environment of current research information systems (CRIS systems). On some typical examples taken from the Slovenian CRIS system SICRIS it was demonstrated that, due to poor access to relevant information, the introduction of ontologies would be sensible primarily in systems that acquire their information from heterogeneous relation databases. Improved access to relevant information was suggested by means of an upgrade of the information system with a semantic component. A special model of ontological infrastructure was designed, and dedicated test ontology was authored by using the tool Protege. The work also introduces a new simplified algorithm for creating ontology (entering ontological classes), the basis of which is the distinction between new and already existing classes in terms of content. The ontology authored based on this algorithm is targeted and can be used for purposes of a search engine. Its main advantage is the descriptive procedure of creating ontology, which makes it smaller and in turn simplifies the process of its creation. The goal of both the ontological infrastructure and the algorithm for creating ontology is the most efficient support to the process of searching for relevant results. The support for increasing the relevance of search results in the proposed model is proposed with a faceted way of narrowing down the existing search results using new metadata generated from ontology. With metadata acquired in such a way, greater efficiency when searching for relevant data is expected as search precision is increased within the existing retrieval when using the described method. Finally, a simulation of the search process with the use of simplified ontology of Slovenian research activities was performed for the abovementioned model in Protégé. Informacijski sistemi o raziskovalni dejavnosti CRIS sistemi metapodatkovni modeli ontologije ontološko podprto iskanje fasetirano iskanje Current research information systems CRIS systems metadata models for ontology ontological-based search faceted information search true true false [A. Bošnjak] Slovenski jezik Angleški jezik Doktorska disertacija 2016-09-06 12:48:27 2016-10-10 12:53:38 2022-07-01 03:08:59 2019-07-11 13:17:04 2016 0 Maribor 0 0000-00-00 NiDoloceno NiDoloceno NiDoloceno 0000-00-00 0000-00-00 0000-00-00 004.775:004.78(043.3) 19861014 URN:SI:UM:DK:SKR8OSSW DOK_Bosnjak_Ales_2016.pdf DOK_Bosnjak_Ales_2016.pdf 1 1010B54172A66DA964D399F564D5AC5E b349cf096d01c186ec92763d12c8469c05764bbdc17fc8a0ddd2180b3ca28431 b68ef8de-7c0d-11eb-bb7a-00155d0001ca https://dk.um.si/Dokument.php?lang=slv&id=107003 Fakulteta za elektrotehniko, računalništvo in informatiko 0 0 0