1. Influence of highly inflected word forms and acoustic background on the robustness of automatic speech recognition for human–computer interactionAndrej Žgank, 2022, izvirni znanstveni članek Opis: Automatic speech recognition is essential for establishing natural communication with
a human–computer interface. Speech recognition accuracy strongly depends on the complexity of
language. Highly inflected word forms are a type of unit present in some languages. The acoustic
background presents an additional important degradation factor influencing speech recognition
accuracy. While the acoustic background has been studied extensively, the highly inflected word
forms and their combined influence still present a major research challenge. Thus, a novel type of
analysis is proposed, where a dedicated speech database comprised solely of highly inflected word
forms is constructed and used for tests. Dedicated test sets with various acoustic backgrounds were
generated and evaluated with the Slovenian UMB BN speech recognition system. The baseline word
accuracy of 93.88% and 98.53% was reduced to as low as 23.58% and 15.14% for the various acoustic
backgrounds. The analysis shows that the word accuracy degradation depends on and changes
with the acoustic background type and level. The highly inflected word forms’ test sets without
background decreased word accuracy from 93.3% to only 63.3% in the worst case. The impact of
highly inflected word forms on speech recognition accuracy was reduced with the increased levels of
acoustic background and was, in these cases, similar to the non-highly inflected test sets. The results
indicate that alternative methods in constructing speech databases, particularly for low-resourced
Slovenian language, could be beneficial.
Ključne besede: human–computer interaction, automatic speech recognition, acoustic modeling, highly inflected word forms, acoustic background
Objavljeno v DKUM: 28.03.2025; Ogledov: 0; Prenosov: 2
 Celotno besedilo (1,12 MB)
Gradivo ima več datotek! Več... |
2. Models of privacy and disclosure on social networking sites: a systematic literature reviewLili Nemec Zlatolas, Luka Hrgarek, Tatjana Welzer-Družovec, Marko Hölbl, 2022, izvirni znanstveni članek Opis: Social networking sites (SNSs) are used widely, raising new issues in terms of privacy and
disclosure. Although users are often concerned about their privacy, they often publish information
on social networking sites willingly. Due to the growing number of users of social networking sites,
substantial research has been conducted in recent years. In this paper, we conducted a systematic
review of papers that included structural equations models (SEM), or other statistical models with
privacy and disclosure constructs. A total of 98 such papers were found and included in the analysis.
In this paper, we evaluated the presentation of results of the models containing privacy and disclosure
constructs. We carried out an analysis of which background theories are used in such studies and
have also found that the studies have not been carried out worldwide. Extending the research to
other countries could help with better user awareness of the privacy and self-disclosure of users
on SNSs.
Ključne besede: structural equations modeling, social networking sites, privacy, disclosure
Objavljeno v DKUM: 28.03.2025; Ogledov: 0; Prenosov: 2
Celotno besedilo (2,50 MB)
Gradivo ima več datotek! Več... |
3. Evolution of domain-specific modeling language: an example of an industrial case study on an RT-sequencerTomaž Kos, Marjan Mernik, Tomaž Kosar, 2022, izvirni znanstveni članek Opis: Model-driven engineering is a well-established software development methodology that
uses models to develop applications where the end-users with visual elements model abstractions
from a specific domain. These models are based on domain-specific modeling language (DSML),
which is particular to the problem domain. During DSML use, new ideas emerge and DSMLs evolve.
However, reports on DSML evolution are rare. This study presents a new DSML called RT-Sequencer
that evolved from our DSML Sequencer to support, in addition to the Data Acquisition domain,
also a new domain—Real-Time Control (RTC) systems. The process of defining models with a new
language RT-Sequencer has changed in a way that new end-users were introduced—advanced endusers, which use general-purpose language (GPL) and advanced programming concepts to define
modeling environments for the RT-Sequencer end-users. More specifically, an industrial experience
with the RT-Sequencer is presented, where DSML was opened for extension so that a GPL code
could be inserted into the model to create new visual blocks for the end-user, and the possibility to
adapt and optimize the execution code for a particular task. Our experience shows the specific case
of DSML evolution supporting another problem domain, and the implementation effort needed to
extend domain-specific modeling language with GPL support.
Ključne besede: model-driven engineering, domain-specific modeling languages, measurement systems, Real-Time Control systems, data acquisition, language evolution, experience report
Objavljeno v DKUM: 27.03.2025; Ogledov: 0; Prenosov: 2
Celotno besedilo (1,70 MB)
Gradivo ima več datotek! Več... |
4. Modeling of tensile test results for low alloy steels by linear regression and genetic programming taking into account the non-metallic inclusionsMiha Kovačič, Uroš Župerl, 2022, izvirni znanstveni članek Opis: Štore Steel Ltd. is one of the biggest flat spring steel producers in Europe. The main
motive for this study was to study the influences of non-metallic inclusions on mechanical properties
obtained by tensile testing. From January 2016 to December 2021, all available tensile strength data
(472 cases–472 test pieces) of 17 low alloy steel grades, which were ordered and used by the final
user in rolled condition, were gathered. Based on the geometry of rolled bars, selected chemical
composition, and average size of worst fields non-metallic inclusions (sulfur, silicate, aluminium
and globular oxides), determined based on ASTM E45, several models for tensile strength, yield
strength, percentage elongation, and percentage reduction area were obtained using linear regression
and genetic programming. Based on modeling results in the period from January 2022 to April 2022,
five successively cast batches of 30MnVS6 were produced with a statistically significant reduction
of content of silicon (t-test, p < 0.05). The content of silicate type of inclusions, yield, and tensile
strength also changed statistically significantly (t-test, p < 0.05). The average yield and tensile strength
increased from 458.5 MPa to 525.4 MPa and from 672.7 MPa to 754.0 MPa, respectively. It is necessary
to emphasize that there were no statistically significant changes in other monitored parameters.
Ključne besede: mechanical properties, tensile test, tensile strength, yield strength, percentage elongation, percentage reduction area, low alloy steel, modeling, linear regression, genetic programming, industrial study, steel making, optimization
Objavljeno v DKUM: 24.03.2025; Ogledov: 0; Prenosov: 2
Celotno besedilo (3,72 MB)
Gradivo ima več datotek! Več... |
5. Dissimilarity-driven behavior and cooperation in the spatial public goods gameYinhai Fang, Tina Perc Benko, Matjaž Perc, Haiyan Xu, 2019, izvirni znanstveni članek Opis: In this paper, we explore the impact of four different types of dissimilarity-driven behavior on the evolution of cooperation in the spatial public goods game. While it is commonly assumed that individuals adapt their strategy by imitating one of their more successful neighbors, in reality only very few will be awarded the highest payoffs. Many have equity or equality preferences, and they have to make do with an average or even with a low payoff. To account for this, we divide the population into two categories. One consists of payoff-driven players, while the other consists of dissimilarity-driven players. The later imitate the minority strategy in their group based on four different dissimilaritydriven behaviors. The rule that most effectively promotes cooperation, and this regardless of the multiplication factor of the public goods game, is when individuals adopt the minority strategy only when their payoff is better than that of their neighbors. If the dissimilarity-driven players adopt the minority strategy regardless of the payoffs of others, or if their payoff is the same, the population typically evolves towards a neutral state where cooperators and defectors are equally common. This may be beneficial when the multiplication factor is low, when defectors would otherwise dominate. However, if the dissimilarity-driven players adopt the minority strategy only when their payoff is worse than that of their neighbors, then cooperation is not promoted at all in comparison to the baseline case in the absence of dissimilarity-driven behavior. We explore the pattern formation behind these results, and we discuss their wider implications for the better understanding of cooperative behavior in social groups.
Ključne besede: theoretical biology, evolution, agent-based modeling, complex system, network science, evolutionary game theory
Objavljeno v DKUM: 26.02.2025; Ogledov: 0; Prenosov: 3
Celotno besedilo (5,13 MB)
Gradivo ima več datotek! Več... |
6. Identification of influential invaders in evolutionary populationsGuoli Yang, Tina Perc Benko, Matteo Cavaliere, Jincai Huang, Matjaž Perc, 2019, izvirni znanstveni članek Opis: The identification of the most influential nodes has been a vibrant subject of research across the whole of network science. Here we map this problem to structured evolutionary populations, where strategies and the interaction network are both subject to change over time based on social inheritance. We study cooperative communities, which cheaters can invade because they avoid the cost of contributions that are associated with cooperation. The question that we seek to answer is at which nodes cheaters invade most successfully. We propose the weighted degree decomposition to identify and rank the most influential invaders. More specifically, we distinguish two kinds of ranking based on the weighted degree decomposition. We show that a ranking strategy based on negative-weighted degree allows to successfully identify the most influential invaders in the case of weak selection, while a ranking strategy based on positive-weighted degree performs better when the selection is strong. Our research thus reveals how to identify the most influential invaders based on statistical measures in dynamically evolving cooperative communities.
Ključne besede: theoretical biology, evolution, agent-based modeling, complex system, network science, evolutionary game theory
Objavljeno v DKUM: 26.02.2025; Ogledov: 0; Prenosov: 4
Celotno besedilo (3,95 MB)
Gradivo ima več datotek! Več... |
7. The importance of depropagation in free radical co-polymerization of itaconates : PhD ThesisTomaž Pirman, 2024, doktorska disertacija Opis: Biobased monomers offer a means to enhance the sustainability of acrylic resins, which are traditionally manufactured from acrylate and methacrylate monomers derived from petrochemical sources. However, utilizing renewable feedstocks for these monomers is currently not economically viable, necessitating the exploration of alternative pathways for producing sustainable acrylic resins. Itaconates, which are produced via fermentation, share a similar structure with conventional monomers, making it possible to achieve comparable polymer properties without significantly changing production processes. However, itaconates have limitations: they depropagate at lower temperatures and exhibit much slower polymerization rates than conventional methacrylates during homopolymerization. This study explores the copolymerization of itaconates with acrylates under standard commercial conditions as a promising approach to address these challenges.
The copolymerization of dibutyl itaconate (DBI) with butyl acrylate (BA) is examined in detail. Initial investigations involved batch homopolymerizations of DBI, varying monomer and initiator concentrations, solvent types, and reaction temperatures to validate known propagation and termination kinetics and to gain a deeper understanding of DBI depropagation. Conversion rates and molecular weight distributions were measured using 1H NMR and size exclusion chromatography, respectively. These results were used to refine the representation of depropagation kinetics by fitting the data to a comprehensive kinetic model of DBI homopolymerization, implemented in the PREDICI software, with additional kinetic parameters sourced from recent literature.
Subsequently, BA was employed as a highly reactive, non-depropagating comonomer in a batch system to mitigate the strong depropagation and slow polymerization rates of DBI. Experiments were conducted with varying initial monomer concentrations and compositions at different temperatures. In addition to demonstrating that BA significantly enhances the polymerization rate, the data provided initial estimates of reactivity ratios for acrylate-itaconate copolymerization.
Furthermore, semi-batch copolymerization experiments were performed with varying monomer compositions under starved feed conditions and elevated temperatures, reflecting typical commercial practices. These trials indicated that resins containing up to 50% itaconate are feasible for production, offering the potential to greatly increase the sustainable content of acrylic polymer resins.
Finally, a comprehensive kinetic model for BA/DBI copolymerization, incorporating itaconate depropagation and acrylate backbiting, was developed using the PREDICI software. The model's predictions are compared to experimental batch and semi-batch conversion profiles, polymer molar mass distributions, and comonomer composition drift profiles. This comparison enhances the understanding of the system and provides a tool for further process development and optimization of operating conditions.
Ključne besede: Itaconates, Bio-based monomer, Kinetic modeling, Sustainable acrylic resins, Depropagation
Objavljeno v DKUM: 12.02.2025; Ogledov: 0; Prenosov: 11
Celotno besedilo (5,03 MB) |
8. Enzyme cascade to enzyme complex phase-transition-like transformation studied by the maximum entropy production principleAndrej Dobovišek, Tina Blaževič, Samo Kralj, Aleš Fajmut, 2025, izvirni znanstveni članek Opis: In biological cells, soluble enzymes often spontaneously reorganize into higher-order complexes called metabolons, providing regulatory advantages over individual soluble enzymes under specific conditions. Despite their importance, the mechanisms underlying metabolon formation remain unclear. Here we report a theoretical model that elucidates the spontaneous transition between soluble enzyme cascades and complexes, driven by fluctuations in intermediate metabolite concentrations. The model integrates the maximum entropy production principle (MEPP) and the Shannon information entropy (MaxEnt), Landau phase-transition theory, kinetic modeling, stability analysis, and metabolic control analysis. Our results show that soluble enzymes and enzyme complexes represent two distinct catalytic states with unique kinetic and regulatory properties. The transition from an enzyme cascade to an enzyme complex displays features of a first-order phasetransition, highlighting the system's tendency to reorganize into its most thermodynamically favorable state, providing a potential pathway for metabolic regulation.
Ključne besede: theoretical modeling, irreversible thermodynamics, maximum entropy production principle, Shannon information entropy, first-order phase transition, enzyme organization, enzyme cascade, enzyme complex
Objavljeno v DKUM: 06.02.2025; Ogledov: 0; Prenosov: 1
Celotno besedilo (1,52 MB)
Gradivo ima več datotek! Več... |
9. |
10. The appropriate work environment for older employees : the case of SloveniaMaja Rožman, Vesna Čančer, 2021, izvirni znanstveni članek Opis: Background: The increase in population life expectancy in developed economies is also reflected in the aging of the workforce; therefore, enterprises should create an appropriate work environment for all employees, emphasizing the older ones.
Objectives: The major objectives of this paper are to establish the impact of appropriate working conditions and training programs for older employees on their work motivation, as well as to establish the latter’s impact on the older employees’ work engagement in Slovenian medium-sized and large enterprises.
Methods/Approach: Structural equation modeling explored the links between four constructs – appropriate working conditions, training programs, work motivation, and work engagement.
Results: In Slovenian enterprises, appropriate working conditions and training programs for older employees positively impact their work motivation. Moreover, the work motivation of older employees positively impacts their work engagement.
Conclusions: These findings can be useful for employers or managers for developing targeted employee motivation and employee engagement programs to leverage the talents and dedication of older employees.
Ključne besede: appropriate work environment, older employees, enterprises, structural equation modeling
Objavljeno v DKUM: 09.12.2024; Ogledov: 0; Prenosov: 12
Celotno besedilo (392,56 KB)
Gradivo ima več datotek! Več... |
