University of Niš: Facta Universitatis (E-Journals) / Универзитет у Нишу
Not a member yet
4905 research outputs found
Sort by
TRENDOVI PRIMENE AI I KOMAPRATIVNA PREDNOST U EVROPSKIM ZEMLJAMA PRE 2020. GODINE
No full textRapid advances in technology, especially AI applications, have contributed to economic transformation at both the microeconomic and macroeconomic levels. Given its substantial potential, AI adoption can become an important source of a country’s comparative advantage. This study examines whether European countries achieved a comparative advantage derived from AI applications during the period 2014–2019. Comparative advantage is assessed using the RSCA index, where positive values indicate its presence. The index is calculated based on the AI patent applications per one million people. The findings reveal that only a limited number of European countries possess a comparative advantage in AI, and that these countries are at varying levels of economic development. This suggests that a nation’s level of economic development is not a decisive factor for the emergence of innovative ideas driven by AI adoption.Napredne tehnologije, a naročito primena AI se razvijaju velikom brzinom i doprinose reformi privrede na mikro i makro nivou. Istovremeno, s obzirom na veliki potencijal koji ima, primena AI može postati značajan izvor komparativne prednosti zemlje. U radu se istražuje da li su evropske zemlje dostigle komparativnu prednost koja potiče iz primene AI u periodu od 2014. do 2019. godine. Komparativna prednost je merena pomoću RSCA indeksa čije pozitivne vrednosti ukazuju na njeno postojanje. Parametar koji je korišćen za izračunavanje ovog indeksa je broj prijavljenih patenata u vezi primene AI. Rezultati su pokazali da mali broj evropskih zemalja ima komparativnu prednost zasnovanu na AI, ali da su u toj grupi zemlje na različitom nivou ekonomskog razvoja. To upućuje na zaključak da razvijenost zemlje nije ključna za razvijanje inovativnih ideja zasnovanih na primeni AI
OBRASCI KORUPCIJE I DEMOKRATSKA KONSOLIDACIJA U SRBIJI
No full textThis paper examines the patterns of corruption and their role in the consolidation of democracy in Serbia after 2000. We begin by assessing the mechanisms by which corruption undermines the fragile balance between institutions, norms and rules that provide trust and legitimacy of the system, thus affecting democratic backsliding. Employing Wolfgang Merkel’s model of embedded democracy as an analytical framework, and drawing on quantitative data from Transparency International’s Corruption Perceptions Index (CPI) and the V-Dem dataset covering the period 2000–2024, the paper explores the mechanisms through which corruption exacerbates democratic backsliding. We then shift our focus towards an examination of the patterns of corruption and democratic backsliding in Serbia. By using data from Transparency International and V-dem database, the paper identifies key forms of corruption and their consequences for democratic processes in Serbia. The findings indicate that corruption is deeply entrenched in state institutions and the political system. Since lack of political will and weak implementation of reforms slow down democratic consolidation, greater transparency, strengthening of independent institutions and active involvement of citizens are necessary for the consolidation of democracy in Serbia.Ovaj rad analizira obrasce korupcije i njihovu ulogu u konsolidaciji demokratije u Srbiji nakon 2000. godine. Najpre se razmatraju mehanizmi putem kojih korupcija podriva krhku ravnotežu između institucija, normi i pravila koja obezbeđuju poverenje i legitimitet sistema, čime se utiče na eroziju demokratije. Koristeći Merkelov model ukorenjene demokratije, rad istražuje načine na koje korupcija produbljuje krizu demokratije. Zatim se fokus premešta na ispitivanje obrazaca korupcije i demokratskog nazadovanja u Srbiji. Na osnovu podataka organizacije Transparency International i baze V-Dem, rad identifikuje ključne oblike korupcije i njihove posledice po demokratske procese u Srbiji. Nalazi ukazuju na to da je korupcija duboko ukorenjena u državnim institucijama i političkom sistemu. Kako nedostatak političke volje i slaba primena reformi usporavaju demokratsku konsolidaciju, veća transparentnost, jačanje nezavisnih institucija i aktivno uključivanje građana predstavljaju nužne preduslove za konsolidaciju demokratije u Srbiji
HARNESSING DEEP LEARNING FOR LUNG CANCER DETECTION USING CT SCAN IMAGES
Get PDFLung cancer persists to be considered as the primary cause of rising mortality rates worldwide with clinical projection depending upon early-stage detection. Detection at the preliminary stage is crucial for treatment of the disease also to further the prognosis process. In this study IQOTHNCCD dataset is used. An innovative model is constructed for distinguishing images into three categories. The dataset comprises of three types of images such as malignant images, benign images and normal images. Preprocessing of the images is done with intensive steps in order to remove unwanted data. Deep learning along with transfer learning have been applied to the image dataset for classifying the images into various categories. In biomedical image classification deep learning serves as the most effective technique for classification and detecting the abnormal pulmonary nodules. Techniques such as InceptionV3, ResNet50, VGG16, VGG19, MobileNetV2 are evaluated through experiment for ensuring the credibility of the designed model. Key performance indicators were used which includes accuracy, sensitivity, precision and F1 score. The results obtained through the proposed model which is the custom CNN model yielded an accuracy of 98.79%, sensitivity of 98.97%, precision of 96.6% and F1 score of 97.7%
MICROWAVE SENSOR ARRAY BASED ON SPLIT RING RESONATORS FOR TUMOR DETECTION AND LOCALIZATION
Get PDFA microwave sensor array for tumor detection and localization was designed and fabricated on an FR‑4 substrate. The sensor consists of three square-shaped split‑ring resonators coupled to a microstrip feeding line. Each sensing cell produces a stopband in the transmission spectrum of the device. The sensing cells are mutually decoupled, so that the deposition of the sample on one of the cells induces shift of resonant frequency of only that cell, thereby enabling spatial resolution. The sensing performance was experimentally verified with the samples of animal tissues in an ex vivo setting. Pork fat and lean tissues were used for preparing the samples, since the dielectric contrast between them roughly corresponds to the contrast between healthy and tumorous tissue. Experimental results suggest that the designed sensor can be effectively used for tumor detection and localization
PRIMENA TEORIJE RELEVANCIJE U ANALIZI MULTIMODALNE KOMUNIKACIJE
No full textThe paper examines the interpretation procedure of multimodal messages from a relevance-theoretic perspective. The aim is to test the potential of the RT apparatus to various types of multimodal messages – advertisements, political posters and memes. The chosen modes of communication have the objective of providing genre diversity for the sample analysis conducted in this paper. Our analysis has the following specific goals: 1) analyzing the explicit content of the message (visual and textual, when applicable), 2) analyzing the implicit content of the message (including possible weak implicatures), 3) potential differences in interpretation that are related to cognitive environment in which they are processed and 4) how the informative and communicative intention can be achieved. The results confirm that RT is a well-grounded framework for future research regarding multimodal communication.U radu se ispituje interpretacija multimodalnih poruka iz pozicije teorije relevancije. Cilj je testirati mogućnosti kognitivno-inferencijalnog aparata kada su u pitanju različiti tipovi multimodalnih dokumenata – reklamni sadržaj, politički posteri i mimovi. Odabrani modaliteti komunikacije imaju za cilj da obezbede žanrovsku raznovrsnost u okviru uzorka analiziranog u ovom radu. Analiza ima sledeće posebne ciljeve: 1) analiza eksplicitnog sadržaja poruke (vizuelnog i tekstualnog), 2) analiza implicitnog sadržaja poruke (uključujući i potencijalne slabe implikature), 3) moguće razlike u interpretaciji koje se odnose na kognitivno okruženje u kojem se poruke interpretiraju i 4) način na koji se mogu ostvariti informativna i komunikativna namera. Rezultati potvrđuju da je teorija relevancije dobro utemeljen okvir za buduća istraživanja multimodalne komunikacije
PREVOD NEPREVODIVOG: ADAPTACIJA U PREVOĐENJU ADMINISTRATIVNIH I PRAVNIH TEKSTOVA NA PRIMERIMA FRANCUSKIH I SRPSKIH DOKUMENATA
No full textCet article vise à identifier et examiner les difficultés dans le processus de traduction de documents administratifs et juridiques du français vers le serbe et vice versa, notamment celles qui concernent les concepts nouveaux ou inexistants dans la langue cible. Notre analyse est fondée sur la théorie de la pertinence (Sperber & Wilson 1989) et s’appuie sur un corpus constitué d’exemples issus des traductions officielles de documents français et serbes. La technique de traduction la plus utilisée est l’adaptation qui s’impose comme nécessaire pour rendre la traduction compréhensible au destinataire. Il existe également une corrélation entre le degré de différences organisationnelles entre les systèmes des deux pays et la traduction de certains termes : plus les différences dans l’organisation des systèmes administratifs et juridiques sont grandes, plus la technique d’adaptation est présente.U radu se razmatraju poteškoće i izazovi u prevođenju administrativnih i pravnih tekstova sa francuskog na srpski jezik i obrnuto, naročito primeri koji se odnose na pojmove koji ne postoje u ciljnom jeziku. Naša analiza, zasnovana na teoriji relevantnosti i sprovedena na primerima iz zvaničnih prevoda različitih dokumenata iz oblasti administracije i prava, obuhvatila je prenošenje imena fizičkih lica, kao i prevođenje naziva institucija, naziva dokumenata i drugih termina. Rezultati istraživanja su pokazali da je najzastupljenija prevodilačka tehnika adaptacija. Takođe, uočena je korelacija između stepena razlika u organizaciji administrativnog i pravosudnog sistema Francuske i Srbije i prevođenja određenih termina: što su navedene razlike između ovih sistema veće, adaptacija je zastupljenija
O PREDLOGU NA U SRPSKOM JEZIKU U KONSTRUKCIJAMA ZA IZRAŽAVANJE NAČINA I O NJEGOVIM FRANCUSKIM EKVIVALENTIMA
No full textLe travail que nous présentons traite de la préposition na en serbe dans les constructions exprimant la manière de déroulement d’un évènement et de ses équivalents français. Notre but est de trouver et distinguer cet usage et d’éclairer la raison pour laquelle certaines prépositions françaises lui correspondent sémantiquement. Plus précisément, la préposition serbe na peut exprimer la manière dans l’interaction des éléments qui se combinent avec elle. En fait, le contexte dans lequel figure cette préposition est ce qui est le plus important dans ces cas. Nous allons présenter une analyse contrastive des exemples serbes et de leurs équivalents français afin de montrer de quelle manière et avec quelle préposition française, chargés de certains arguments, on peut exprimer la façon de déroulement d’une action.Ovaj rad proučava predlog na u srpskom jeziku u konstrukcijama sa značenjem priloške odredbe za način. Težićemo da otkrijemo njegove prevodne ekvivalente u francuskom jeziku. Pritom, potrebno je istaći elemente od kojih zavisi ispoljavanje određene značenjske vrednosti predloga na u rečenici. Cilj nam je da ukažemo i na predloge koji se u francuskom jeziku koriste da iskažu ovakvo značenje i da ispitamo da li se semantički podudaraju sa srpskim predlogom na. Treba naglasiti da je kontekst presudan za otkrivanje značenja koje ispoljava predlog na. Kontrastivna analiza srpskih primera i njihovih francuskih ekvivalenata utvrdila je da se u konstrukcijama sa takvim značenjem najčešće koriste francuski predlozi de i à, a ređe polivalentna reč comme
O PERCEPCIJI U FRANCUSKOJ SREDNJOVEKOVNOJ KNJIŽEVNOSTI: PRIKAZ PET ČULA U DIDAKTIČKIM SPISIMA
No full textLe présent article examine la présence des sens et de la perception dans les textes didactiques français en langue vulgaire, datant du XIIIe siècle : le Bestiaire du pseudo-Pierre de Beauvais, Le Bestiaire d’Amour de Richard de Fournival, l’Image du monde de Gossouin de Metz et Placides et Timéo de l’auteur anonyme. Étant donné que la littérature didactique puisait dans les sources antiques et bibliques, le corpus examiné manifeste clairement ces traits ; toutefois, la disparition de l’allégorie des textes didactiques, qui s’est initiée à cette époque, a laissé également les traces. Sans être le sujet à part dans le corpus analysé, les sens et la perception réapparaissent comme thèmes, liés surtout à l’acquisition du savoir.U radu je analiziran prikaz pet čula i značaja percepcije u delima francuske srednjovekovne didaktičke književnosti iz XIII veka: Bestijarijumu pseudo-Pjera de Bovea, Ljubavnom bestijarijumu Rišara de Furnivala i dva enciklopedijska dela (Slika sveta Gosuena iz Meca i Plasid i Timeo nepoznatog autora). U obrađenom korpusu čula su prikazana sporadično, u kontekstu prikaza životinja koje su u srednjovekovnoj književnosti bile simboli čula, potom kao način sticanja znanja; u Ljubavnom bestijarijumu kurtoazna potraga, ilustrovana simbolima životinja, organizovana je delom i u skladu sa čulima. U obrađenom korpusu, iako se uočava opadanje značaja simbolike religiozne motivacije, ipak nije prisutan koherentni naučni pristup u prikazu čula i percepcije, što je u skladu sa kulturom date epohe
VIBRATION MITIGATION-BASED MACHINE LEARNING-DRIVEN DESIGN OF METASTRUCTURES
Get PDFThis research is concerned with the development of a longitudinally excited metastructure, featuring periodically distributed external units, each equipped with internal oscillators functioning as vibration absorbers. Initially, the metastructure designed for vibration attenuation around the first structural resonance, is characterized by uniformity, with all absorbers being identical and consisting of cantilevers integrated into the external components, each cantilever terminating in a concentrated mass block. This study employs a machine learning approach to maximize vibration attenuation efficiency around the second resonance, as well as concurrently at the first and second resonant frequencies in two associated optimality criteria related to the width of the attenuation region and the amplitude reduction, respectively. The new metastructures redesigned based on these criteria are fabricated by 3D printing, and their enhanced vibration mitigation capabilities are verified experimentally
FINITE ELEMENT MODELLING OF IMPEDANCE TUBE TEST
Get PDFImpedance tubes are used to characterize porous acoustic materials. The report proposes a suitable model for numerical simulations of the test of materials in an impedance tube. Numerical simulations make it possible to generate synthetic data on the acoustic characteristics of the material when varying the parameters in the material model. These data are suitable for using machine learning algorithms as well as for solving inverse problems for identifying the material parameters in the mathematical model. The Johnson-Champoux-Allard model is used in the work, which is suitable for a wide class of porous acoustic materials