Digital library of University of Maribor
Not a member yet
81039 research outputs found
Sort by
Impact of artificial intelligence on employment
Get PDFUmetna inteligenca (UI) je ena ključnih tehnologij četrte industrijske revolucije, ki že povzroča pomembne spremembe v družbi in na trgu dela. Magistrska naloga obravnava vpliv UI na zaposlovanje ter preoblikovanje delovnih okolij v različnih gospodarskih sektorjih. Raziskava se osredotoča na proizvodni, storitveni in razvojni sektor, pri čemer analiziramo, kako zaposleni in vodstvo zaznavajo prisotnost UI v organizacijah ter kakšne izzive in priložnosti prinaša njena implementacija. Cilj raziskave je ugotoviti, v kolikšni meri so zaposleni in vodstvo seznanjeni z vplivom tehnologije na delovna mesta ter kako se kadrovski management sooča s spremembami v zaposlovanju.
V teoretičnem delu smo predstavili zgodovinski razvoj tehnologij, opredelili ključne inovacije in preučili sodobne trende v zaposlovanju. Poseben poudarek smo namenili kadrovskemu managementu v kontekstu UI, kjer smo analizirali vpliv tehnologije na procese zaposlovanja, izbire kandidatov in upravljanje talentov. Empirični del temelji na kombinaciji kvantitativnih in kvalitativnih metod. Izvedli smo anketno raziskavo ter opravili intervjuja s predstavnikoma izbranih podjetij.
Rezultati raziskave nakazujejo, da je dejanska uporaba UI v trenutnem delovnem okolju še vedno omejena. Vodstva organizacij sicer prepoznavajo njen potencial, vendar niso v celoti pripravljena na spremembe. Ključni izzivi, s katerimi se soočajo, vključujejo etične dileme, varnost podatkov in transparentnost algoritmov pri izbiri kandidatov. Prav tako še niso vzpostavljene ustrezne strategije za krepitev medgeneracijske pomoči pri usposabljanju zaposlenih za delo z UI. Za uspešno implementacijo so ključni vlaganje v usposabljanje zaposlenih, razvoj organizacijske kulture ter prilagoditev regulativnih okvirjev, ki bodo zagotavljali etično in pravično uporabo UI v procesih zaposlovanja.Artificial intelligence (AI) is one of the key technologies of the Fourth Industrial Revolution, already driving significant changes in society and the labor market. This master\u27s thesis examines the impact of AI on employment and the transformation of work environments across various economic sectors. The research focuses on the manufacturing, service, and development sectors, analyzing how employees and management perceive the presence of AI in organizations and what challenges and opportunities its implementation presents. Objective is to determine the extent to which employees and management are aware of AI’s impact on jobs and how HRM is addressing changes in employment.
Theoretical part provides a historical overview of technological development, defines key innovations, and explores employment trends. Special emphasis is placed on HRM in the context of AI, analyzing its influence on hiring processes, candidate selection, and talent management. Empirical part is based on a combination of quantitative and qualitative methods. A survey was conducted with representatives from two selected companies.
Findings indicate that the actual use of AI in the current work environment remains limited. While management recognizes its potential, organizations are not yet fully prepared for these changes. Key challenges include ethical dilemmas, data security, and algorithm transparency in candidate selection. Additionally, structured strategies for strengthening intergenerational support in AI-related employee training have not yet been established. For successful implementation, it is essential to invest in employee training, develop organizational culture, and adapt regulatory frameworks to ensure the ethical and fair use of AI in recruitment and employment processes
Material model of composite structure and topology optimisation of composite monocoque
Get PDFMagistrska naloga se osredotoča na razvoj materialnega modela kompozitne strukture in optimizacijo topologije karbonskega monokoka za dirkalnik ekipe Formula Student, UNI Maribor Grand Prix Engineering. Cilj naloge je pridobiti natančne podatke o mehanskih lastnostih kompozitne strukture, ki bodo uporabljeni v numeričnih simulacijah z metodo končnih elementov v programu Ansys. Na podlagi teh podatkov bo izvedena osnovna topološka optimizacija monokoka, pri čemer bo glavni poudarek na zmanjšanju mase ob ohranjanju strukturne trdnosti.
V okviru raziskave bo izvedena analiza obstoječih materialnih modelov, optimizacija geometrije in strukture monokoka ter validacija rezultatov s pomočjo eksperimentalnih podatkov. Glavne predpostavke vključujejo obravnavo karbonskih vlaken kot linearno elastičnega materiala v določenem območju obremenitev ter homogeno obravnavo monokoka, čeprav je sestavljen iz več slojev. Omejitve raziskave zajemajo osredotočenost na statične in dinamične obremenitve brez upoštevanja toplotnih vplivov ter prilagajanje optimizacije v skladu s tekmovalnimi predpisi.
Rezultati naloge bodo prispevali k izboljšanju procesov načrtovanja kompozitnih struktur ter omogočili nadaljnji razvoj lahkih in zmogljivih monokokov v okviru ekipe Formula Student.This master\u27s thesis focuses on the development of a material model for a composite structure and the topology optimization of a carbon monocoque for the Formula Student team, UNI Maribor Grand Prix Engineering. The objective is to obtain precise data on the mechanical properties of the composite structure, which will be used in finite element method (FEM) simulations within the Ansys software. Based on this data, an initial topology optimization of the monocoque will be conducted, with the primary goal of reducing weight while maintaining structural integrity.
The research includes an analysis of existing material models, optimization of the monocoque\u27s geometry and structure, and validation of results using experimental data. The key assumptions consider carbon fibers as a linear elastic material within a specific load range and treat the monocoque as a homogeneous structure, despite being composed of multiple layers. Research limitations include a focus on static and dynamic loads without considering thermal effects, as well as adaptation of the optimization process in compliance with competition regulations.
The findings of this thesis will contribute to improving the design processes of composite structures and enable the further development of lightweight and high-performance monocoques within the Formula Student team
Automation of a transport search using web scraping
Get PDFV diplomskem delu smo razvili spletno aplikacijo, ki uporabnikom omogoča enostavno iskanje voznih redov različnih prevoznikov v Sloveniji. Glavni poudarek je bil na avtomatizaciji iskanja s pomočjo spletnega strganja in povezovanju podatkov iz različnih virov v enoten sistem. Implementirali smo podporo za več ponudnikov, geolokacijsko iskanje bližnjih postaj ter predpomnjenje rezultatov.In our thesis, we developed a web application that allows users to search for timetables of various transport providers in Slovenia easily. The main focus was on automating the search using web scraping and integrating data from different sources into a unified system. We implemented support for multiple providers, geolocation-based search for nearby stations, and caching of results
Relations between prisoners and prison staff in celje prison
Get PDFOdnosi med zaporskim osebjem in obsojenci so ključnega pomena za doseganje pozitivnih rehabilitacijskih rezultatov ter igrajo ključno vlogo pri delovanju zaporskega sistema. Vsaka skupina zaporskih akterjev ima svoje vloge, cilje in pričakovanja, kar vpliva na kakovost teh odnosov, ki prispevajo k zmanjšanju konfliktov ter izboljšanju možnosti za uspešno reintegracijo obsojencev v družbo. V zaključnem delu smo se osredotočili na dinamiko odnosov med obsojenci in zaporskim osebjem, kjer smo preučevali dejavnike, ki vplivajo na resocializacijo obsojencev. Analizirali smo domačo in tujo literaturo, ki obravnava odnose znotraj zaporskega sistema. V empiričnem delu smo izvedli kvalitativno študijo o odnosih v zaporskem sistemu, v kateri smo izvedli intervjuje z zaporskimi delavci v Zavodu za prestajanje mladoletniškega zapora in kazni zapora Celje. V intervjuvanju je sodelovalo deset zaporskih delavcev, katerim smo zastavili vprašanja, ki so se navezovala na odnose z obsojenci, obravnavo obsojencev in pripravljenost obsojencev za sodelovanje z zaporskimi delavci ter tretmajske programe. Ugotovitve analize intervjujev so pokazale, da so odnosi med zaporskimi delavci in obsojenci v zaporu pozitivni ter korektni. Intervjuvanci so poudarili, da je pomembno, da se zaporsko osebje drži svojih obljub obsojencem, saj to pozitivno vpliva na njihovo resocializacijo in zaupanje v zaporsko osebje. Uspešna rehabilitacija je dosežena, ko se kombinira z dolgoročnimi pristopi, ki vključujejo osebnostno spreminjanje vedenjskih vzorcev in spodbujanje notranje motivacije. Zaporski delavci so prav tako izpostavili, da je pripravljenost obsojencev za sodelovanje v rehabilitacijskih programih pogosto odvisna od kakovosti odnosa z osebjem ter od njihovega občutka varnosti in spoštovanja v zaporskem okolju. Višja stopnja zaupanja pogosto vodi v večjo odprtost in aktivno vključevanje v tretmajske aktivnosti.The relationships between prison staff and prisoners are crucial for achieving positive rehabilitation outcomes and play a key role in the functioning of the prison system. Each group of prison actors has its own roles, goals, and expectations, which influence the quality of these relationships and contribute to reducing conflicts and improving the chances for prisoners\u27 successful reintegration into society. In the concluding section, we focused on the dynamics of relationships between prisoners and prison staff, emphasising the factors that impact the resocialization of prisoners. We analyzed domestic and international literature addressing relationships within the prison system. In the empirical part, we carried out a qualitative study on relationships within the prison system, conducting interviews with prison staff at the Celje Prison. Ten prison staff members participated in the interviews and were asked questions regarding their relationships with prisoners, the treatment of prisoners, and the prisoners\u27 willingness to cooperate. The findings from the analysis of the interviews showed that relationships in the prison are positive and respectful. The interviewees most frequently emphasised the importance of prison staff keeping their promises to prisoners, as this greatly influences their resocialization and builds trust in the prison staff. Successful rehabilitation is best achieved when combined with long-term approaches that include transforming behavioural patterns and fostering intrinsic motivation. Prison staff also pointed out that prisoners\u27 willingness to participate in rehabilitation programs often depends on the quality of their relationship with the staff, as well as their sense of safety and respect within the prison environment. A higher level of trust often leads to greater openness and active engagement in treatment activities
Validation of a Slovenian version of the healthy lifestyle and personal control questionnaire (HLPCQ) for use with patients in family medicine
Get PDFBackground Chronic non-communicable diseases are the world’s leading cause of death and disability. The
emerging field of lifestyle medicine requires equipping healthcare professionals with instruments, knowledge, skills
and competencies. Measuring an individual’s lifestyle with a valid and reliable instrument is the first step in promoting
it. The aim of the study was to validate the Slovenian adaptation of the Healthy Lifestyle and Personal Control
Questionnaire (HLPCQ).
Methods A cross-sectional study was conducted among 666 questionnaire participants, and they were adult
participants (aged 18 and above) from family medicine practices with cardiovascular diseases (CVDs) risk factors (e.g.,
hypertension, high cholesterol) but without a diagnosis of acute CVDs. The questionnaire included demographic
data and anthropological measures and a translated English HLPCQ questionnaire. The instrument was translated
using the forward-backwards translation method. The study was conducted in accordance with the principles of the
World Medical Association Declaration of Helsinki. In addition to assessing the construct validity of the questionnaire,
exploratory and confirmatory factor analyses were used to determine content and face validity, and internal
consistency reliability.
Results The mean age of male participants was 41.34 (± 13.220) years, the mean age of female participants was
40.31 (± 11.905) years. The Cronbach’s alpha was 0.852, and all questionnaire subscales had positive correlations.
Sampling adequacy was confirmed by the Kaiser-Meyer-Olkin (KMO) index (0.851), and Bartlett’s test of sphericity was
significant (χ² = 4647.694, p < 0.001), indicating suitability for Principal Component Analysis (PCA). PCA revealed a fivefactor solution, accounting for 50.67% of the total variance.
Conclusions The most influential factors for a healthy lifestyle were daily routine, healthy dietary choices, avoidance
of harmful dietary habits, organized physical activity, and social and mental balance. The Slovenian version had high
factor validity and reliability. It can be used in Slovenian Community Health Centre to assess an individual’s control over various lifestyle dimensions. The instrument also holds potential for use in public health initiatives, supporting
early identification of lifestyle-related risk factors and promoting preventive care strategies in the primary care setting
Elementi teroristične organizacije v organiziranem kriminalu, primer mehiških kartelov
Get PDFV diplomskem delu so analizirane aktivnosti mehiških kartelov z vidika ugotavljanja, ali pri delovanju organiziranega kriminala lahko zasledimo tudi elemente, značilne za teroristične organizacije. Na primeru skupin organiziranega kriminala z območja Mehike so analizirane predvsem oblike nasilnega vedenja, ki so značilne za te skupine, in ugotovitve, kakšna tveganja predstavljajo za Mehiko, sosednje države, pa tudi za Evropo. Vsaka kriminalna organizacija ima drugačen sistem delovanja, motivacijo in ideologijo. Zato nimamo le enovitega odgovora na to vprašanje. Nekatere skupine mehiškega organiziranega kriminala so bližje tradicionalnim terorističnim skupinam, medtem ko druge skupine bolj spadajo v kategorijo klasičnega organiziranega kriminala. Razne ugrabitve in umori tujih državljanov v Mehiki, sploh ZDA, vedno sprožijo ukrepanje članov politike in kongresa. Uradno označevanje mehiških kartelov kot teroristične organizacije naj bi imelo več prednosti kot slabosti. Omogoča večji pregon posameznikov, poveča se možnost prepovedi finančnih transakcij in zamrznitev premoženja in omogoča prepoved vstopa v določene države kot tudi odstranitev oseb iz držav in druge. Imenovanje mehiških kartelov kot teroristične organizacije naj bi za ameriške politike spremenilo igro. Politiki menijo, da bodo karteli končno postali tarča in da bodo preganjani tisti, ki jim nudijo finančno podporo, vključno s kitajskimi entitetami, ki jim pošiljajo kemikalije za proizvodnjo drog. Pravzaprav bi države dobile več mrež in orodij za zapiranje kartelov in mrež, ki jih podpirajo.This paper analyses activities of Mexican cartels from the perspective of determining whether elements typical of terrorist organizations can also be identified in organized crime operations. The analysis focuses on forms of violent behavior typical of these groups, using examples of organized crime groups from Mexico, and conclusions about the risks they pose for Mexico, neighboring countries, and Europe. Each criminal organization has a different mode of operation, motivation, and ideology. Therefore, there is no single answer to this question. Some groups of Mexican organized crime are closer to traditional terrorist groups, while others fall more into the category of classic organized crime. Various kidnappings and murders of foreign citizens in Mexico, especially from the US, always prompt action from politicians and Congress members. Officially designating Mexican cartels as terrorist organizations is said to offer more advantages than disadvantages. It allows for greater prosecution of individuals, increases the possibility of banning financial transactions and freezing assets, and enables the prohibition of entry into certain countries as well as the removal of individuals from countries, among other things. Labelling Mexican cartels as terrorist organizations is seen as a game-changer for US politicians. They believe that the cartels will finally become a target, and those providing financial support, including Chinese entities sending them chemicals for drug production, will be prosecuted. In fact, countries would gain more networks and tools to shut down the cartels and the networks that support them
Critical analysis of contentious methods in lie detection
Get PDFSporne metode ugotavljanja laži, kot so analiza mikroobraznih izrazov, tehnologija možganskih prstnih odtisov (angl. brain fingerprinting), analiza očesnih gibov in termografsko zaznavanje, pogosto temeljijo na pomanjkljivo preverjenih znanstvenih predpostavkah in neenotnih interpretacijah. Njihova uporaba v sodnih postopkih sproža številna vprašanja glede zanesljivosti, veljavnosti in etike. V nekaterih primerih, kot je Harrington v. State of Iowa (2003), so bili dokazi, pridobljeni s pomočjo takšnih tehnik, označeni kot nerelevantni ali nezadostno utemeljeni, kar potrjuje pomisleke o njihovi forenzični vrednosti.
Tehnike pogosto ustvarjajo vtis znanstvene natančnosti, a številne študije kažejo na njihovo nizko ponovljivost in visoko stopnjo subjektivnosti. Poleg znanstvenih dvomov se pojavljajo tudi etične dileme, povezane z invazivnostjo nekaterih metod, zlasti pri uporabi nanotehnologij in biometričnih sistemov, ki lahko posegajo v pravico do zasebnosti in spoštovanje telesne integritete. Zaznavanje laži postane še posebej problematično, ko se rezultati interpretirajo brez jasno določenih standardov ali se uporabljajo zunaj sodne dvorane, npr. pri pogajanjih o priznavanju krivde.
Čeprav nobena od analiziranih metod ne zagotavlja nedvoumne in znanstveno potrjene identifikacije laži, se določeni pristopi, kot sta kognitivna analiza izjav in strateška uporaba dokazov, kažejo kot obetavni. Usmerjanje razvoja v smeri metod, ki temeljijo na preverjenih vedenjskih indikatorjih in kognitivnih obremenitvah, ob hkratnem spoštovanju človekovih pravic in strokovne odgovornosti predstavlja možen okvir za prihodnje izboljšave.Contentious methods of lie detection, such as facial microexpression analysis, Brain Fingerprinting, eye movement analysis and thermographic detection, are often based on poorly validated scientific assumptions and inconsistent interpretations. Their application in legal proceedings raises numerous concerns regarding reliability, objectivity, and ethics. In some cases, such as Harrington v. State of Iowa (2003), evidence obtained through these techniques has been deemed irrelevant or insufficiently substantiated, supporting concerns about their forensic value.
These techniques often create the illusion of scientific precision, yet numerous studies indicate low replicability and a high degree of subjectivity. In addition to scientific concerns, ethical dilemmas arise, particularly regarding the invasiveness of certain methods - especially neurotechnologies and biometric systems - which may infringe upon the right to privacy and bodily integrity. Lie detection becomes particularly problematic when results are interpreted without clear standards or used outside the courtroom, for instance, in plea bargaining.
Although none of the analysed methods offers unequivocal and scientifically validated identification of deception, certain approaches - such as cognitive statement analysis and the strategic use of evidence - appear promising. A shift toward methods grounded in verified behavioural indicators and cognitive load principles, while respecting human rights and professional accountability, presents a potential framework for future development
Comparison of generative artificial intelligence models for code generation
Get PDFCilj magistrskega dela je bil primerjalno oceniti kakovost programske kode, ki jo generirajo modeli umetne inteligence ChatGPT (4o, o1), Gemini (Flash, Pro) in Microsoft Copilot. Na področju generativne umetne inteligence in kakovosti programske opreme smo z uporabo kvantitativnih metrik in orodij analizirali kodo, generirano za različno zahtevne naloge. Rezultati kažejo, da vsi modeli ustvarjajo sintaktično pravilno kodo, a se razlikujejo predvsem v funkcionalni pravilnosti, kompleksnosti in berljivosti. Plačljivi modeli so bili pravilnejši, a kompleksnejšibrezplačni (Copilot, Gemini Flash) pa enostavnejši in berljivejši. Priporočamo izbiro modela glede na prioritete projekta.The objective of the master\u27s thesis was to comparatively evaluate the quality of program code generated by the artificial intelligence models ChatGPT (4o, o1), Gemini (Flash, Pro), and Microsoft Copilot. In the field of generative artificial intelligence and software quality, we analyzed code generated for tasks of varying complexity using quantitative metrics and tools. The results show that all models produce syntactically correct code, but differ primarily in functional correctness, complexity, and readability. Paid models were more correct but more complexfree models (Copilot, Gemini Flash) were simpler and more readable. We recommend selecting a model based on project priorities
Comparative study of modern differential evolution algorithms
Get PDFSince the discovery of the Differential Evolution algorithm, new and improved versions have continuously emerged. In this paper, we review selected algorithms based on Differential Evolution that have been proposed in recent years. We examine the mechanisms integrated into them and compare the performance of algorithms. To compare their performances, statistical comparisons were used as they enable us to draw reliable conclusions about the algorithms’ performances. We use the Wilcoxon signed-rank test for pairwise comparisons and the Friedman test for multiple comparisons. Subsequently, the Mann–Whitney U-score test was added. We conducted not only a cumulative analysis of algorithms, but we also focused on their performances regarding the function family (i.e., unimodal, multimodal, hybrid, and composition functions). Experimental results of algorithms were obtained on problems defined for the CEC’24 Special Session and Competition on Single Objective Real Parameter Numerical Optimization. Problem dimensions of 10, 30, 50, and 100 were analyzed. In this paper, we highlight promising mechanisms for further development and improvements based on the study of the selected algorithms
LLM in the loop: a framework for contextualizing counterfactual segment perturbations in point clouds
Get PDFPoint Cloud Data analysis has seen a major leap forward with the introduction of PointNet algorithms, revolutionizing how we process 3D environments. Yet, despite these advancements, key challenges remain, particularly in optimizing segment perturbations to influence model outcomes in a controlled and meaningful way. Traditional methods struggle to generate realistic and contextually appropriate perturbations, limiting their effectiveness in critical applications like autonomous systems and urban planning. This paper takes a bold step by integrating Large Language Models into the counterfactual reasoning process, unlocking a new level of automation and intelligence in segment perturbation. Our approach begins with semantic segmentation, after which LLMs intelligently select optimal replacement segments based on features such as class label, color, area, and height. By leveraging the reasoning capabilities of LLMs, we generate perturbations that are not only computationally efficient but also semantically meaningful. The proposed framework undergoes rigorous evaluation, combining human inspection of LLM-generated suggestions with quantitative analysis of semantic classification model performance across different LLM variants. By bridging the gap between geometric transformations and high-level semantic reasoning, this research redefines how we approach perturbation generation in Point Cloud Data analysis. The results pave the way for more interpretable, adaptable, and intelligent AI-driven solutions, bringing us closer to realworld applications where explainability and robustness are paramount