RFOS - Repository of Faculty of Organizational Sciences Univ. of Belgrade
Not a member yet
2871 research outputs found
Sort by
Brand Management of Urban Tourist Destination Based on Dimensions of Tourist Attractiveness
No full textAt the very core of an urban tourist destination is a multidimensional construct of tourist attractiveness, which should be appealing to potential tourists and influence their decision to visit. Given the challenges of increasing global competition and the negative consequences of the COVID-19
pandemic, urban destinations must adequately identify the key dimensions
of their attractiveness and ensure their visibility and differentiation. The purpose of this paper is to identify the role that different dimensions of tourist
attractiveness have in managing the brand of an urban tourist destination,
with a special focus on forming the expectations, attitudes and intentions of
potential tourists. The methodology used in the paper includes: theoretical
conceptualization of urban tourist destination, definition of dimensions of
tourist attractiveness and determining the role that dimensions of tourist attractiveness have in managing the brand of urban tourist destinations. One
of the conclusions of the research presented in the paper is that the dimensions of tourist attractiveness, both physical and social, play a key role in the
strategic processes of brand management of an urban tourist destination
Tehnologija i društvene nejednakosti
No full textpstrakt: U ovom radu se društvene nejednakosti posmatraju kao nejednak
pristup različitim resursima, posebno onim koji imaju strateški značaj za pojedince i
društvene grupe (društvena moć, materijalno bogatstvo, prihodi, pristup različitim
javnim uslugama, obrazovanje, rad). Savremeni tehnološki razvoj nosi potencijal i za
jačanje i za neutralisanje društvenih nejednakosti. Nasleđena neoliberalna ideologija
je već razvila visok nivo društvenih nejednakosti. Industrijska revolucija treće
generacije, u velikoj meri, isključuje čoveka iz radnog procesa. Ovo proizvodi
opasnost od još veće radikalizacije nejednakosti. Globalna nejednakost, podela na
razvijene i nerazvijene zemlje, se zaoštrava i proizvodi međunarodnu nestabilnost.
Nacionalna društva postaju iznutra duboko podeljena. Sada se, kao razvojni
imperativ, postavlja pitanje socijalnog zbrinjavanja velikog broja suvišnih radnika.
Počinju da se razvijaju strategije kojima bi društvene nejednakosti mogle biti
stavljene pod kontrolu. Cilj ovog rada je da se ukaže na ključne društvene
nejednakosti i moguće puteve njihovog prevazilaženja.
Ključne reči
Framework for evaluation of multimodal biometric systems
No full textPrimena biometrijskih tehnologija danas u ljudskom društvu postaje sve češća, gotovo da možemo konstatovati da je ona deo naše svakodnevnice. Prilikom implementacije biometrijske autentikacije, svaki sistem ima svoje zahteve i ograničenja, u zavisnosti od konkretnog scenarija u kojem se sistem koristi. Za odabir odgovarajućeg biometrijskog modaliteta, kao i algoritama za rad sa biometrijskim modalitetom, neophodno je sprovesti odgovarajuću evaluaciju performansi rada biometrijskog sistema. Ipak, ovu evaluaciju nije uvek lako sprovesti, kako za unimodalne, tako i za multimodalne biometrijske sisteme. Čak i kada su dostupne javne baze biometrijskih podataka za evaluaciju algoritama određenog biometrijskog modaliteta, potrebno je prilagoditi rad sistema protokolu testiranja koji konkretna baza definiše. U slučaju multimodalnog pristupa, evaluacija se dodatno komplikuje usled upotrebe različitih algoritama za fuziju informacija. Kako u dostupnoj relevantnoj literaturi nije pronađen detaljan prikaz modela evaluacije multimodalnih biometrijskih sistema, a radi prevazilaženja ovih teškoća, u okviru ovog doktorata definisan je objedinjeni model evaluacije multimodalnih biometrijskih sistema. Za definisanje ovog modela primenjena je MDA (Model Driven Architecture) paradigma. U okviru objedinjenog modela dat je metamodel evaluacije multimodalnih biometrijskih sistema, koji predstavlja svojevrsnu ontologiju pojmova značajnih za ovu oblast. Primenom ovog metamodela, moguće je kreirati modele evaluacije različitih biometrijskih sistema. Na osnovu modela evaluacije multimodalnih biometrijskih sistema kreiran je prototip okvira za evaluaciju multimodalnih biometrijskih sistema. Pomoću predloženog okvira moguća je evaluacija performansi multimodalnog biometrijskog sistema u različitim slučajevima korišćenja. Eksperimentalni rezultati evaluacije nad konkretnom bazom i algoritmima pokazuju da primena okvira skraćuje za četiri puta vreme potrebno za evaluaciju. Razvijena je i nova metoda za analitičko određivanje praga osetljivosti u skladu sa postavljenim parametrima željenog ponašanja sistema. Na kraju, na primeru alata koji je koristio neke od funkcionalnosti okvira, prikazano je kako primena okvira može učiniti efikasnijim proces obrazovanja inženjera u oblasti biometrije.Application of biometric technologies in our contemporary human society is getting more frequent, so we can almost state that biometric technologies are part of our everyday life. When implementing biometric authentication, each system has specific requirements and constraints, which depend on the actual scenario in which the system is being used. In order to choose the adequate biometric modality, and also a fitting algorithm for the chosen modality, it is necessary to perform an evaluation of the biometric system performance. However, this evaluation is not always easy to conduct. This fact is true for both the unimodal and multimodal biometric systems. Even when open biometrics databases are available for evaluation, it is necessary to adapt system to work with testing protocol of the chosen open database. Moreover, if the biometric system uses multiple biometric modalities, evaluation gets even more complicated because of different available fusion algorithms. In order to overcome these difficulties, as there is not a detailed model of multimodal biometric systems available in relevant literature, this thesis presents a unified multimodal biometric systems evaluation model. Presented model is based on MDA (Model Driven Architecture) paradigm. A part of the unified multimodal biometric systems evaluation model is the metamodel of multimodal biometric system evaluation, which represents an ontology of terms used in this domain. Based on unified multimodal biometric systems evaluation model, a prototype framework for multimodal biometrics systems evaluation has been created. By using proposed framework it is possible to evaluate performance of multimodal biometric system in different use cases. Experimental evaluation results based on used database and algorithms show that the use of framework shortens time necessary for evaluation to a quarter of previously required time. Also, a new analytical method for biometric system threshold optimization, based on the predefined desired system behavior was developed. As final, a learning tool based on some of the framework functionalities is used to show how the use of framework can make the process of educating engineers in the field of biometrics more efficient
Fair classification via Monte Carlo policy gradient method
No full textArtificial intelligence is steadily increasing its impact on everyday life. Therefore, the societal issues of artificial intelligence have become an important concern in the AI research. The presence of data that reflects human biases towards historically discriminated groups defined by sensitive features such as race and gender, results in machine learning models which discriminate against these groups. In order to tackle the impact of bias in data, researchers developed a variety of specialized machine learning algorithms which are able to satisfy different fairness constraints imposed on the model. Group fairness constraints do not fit standard machine learning formulations easily due to their non-differentiable nature. In this paper we developed a technique for learning a fair classifier by Monte Carlo policy gradient method which naturally deals with such non-differentiable constraints. Our methodology focuses on direct optimization of both group fairness metric and predictive performance of the model. In addition, we propose two different variance reduction techniques of gradient estimation. We compare our models to seven other related and state-of-the-art models and demonstrate that they are able to achieve better trade-off between accuracy and unfairness. To the best of our knowledge, this is the first fair classification algorithm which solves the issue of non-differentiable constraints by reinforcement learning techniques
The impact of hybrid workplace models on intangible assets: the case of an emerging country
No full textAs part of a safety-first principle during the COVID-19 pandemics, the vast majority of companies have enabled flexible working environments, reducing the number of employees in the premises. The global best practices have firstly been recorded among the ICT companies which offered teleworking to their employees, empowering safety and flexibility through remote work policies and flexible working hours. Although hybrid working models might become a standard in many industries, only a paucity of papers has examined the relationship between novel working environments and various classes of intangible assets. The aim of this paper is to present the effects of hybrid working models (telework and flexible working hours) on intangible assets (human, relational, structural and intellectual capital). While the existing hybrid work principles have already shown mixed effects on corporate outcomes, its impact on intangible assets remains unrevealed. To address this research gap, we conducted an empirical study. Primary data were collected in the Serbian ICT sector (N=122) using a structured questionnaire developed for this purpose. Data was analyzed with the OLS regression. The results confirm the positive effects of the hybrid working model on intangible assets of ICT companies, which could further propel the financial success of these companies. In general, these results imply that hybrid working models, which are becoming a standard for many industries, would not jeopardize the creation of intangible assets - the ultimate resource of modern companies
Bankruptcy Risk Prediction in Ensuring the Sustainable Operation of Agriculture Companies
No full textIn recent decades, predicting company bankruptcies and financial troubles has become a major concern for various stakeholders. Furthermore, because financially sustainable businesses are affected by numerous highly complex factors, both internal and external, the situation is even more complex. This paper applies Altman's Z-score models; more precisely, the paper applies the initial Z-score model (a model for manufacturing companies), the Z '-score model (for companies operating in emerging markets), and the Z-score bankruptcy probability calculation. Therefore, this paper offers the results of the application of different Z-score models and the calculation of bankruptcy probability on a sample of agricultural companies listed on the Belgrade Stock Exchange in the period 2015-2019. In addition, different Z-score models are used for the same sample so that the difference between their results and application can be determined. In addition, the validity of the data published in the financial statements of the respective companies was confirmed using the Beneish M-score model with five and eight variables. The results obtained by applying Altman's Z-score model (initial and adapted to emerging markets) indicate that a certain number of companies had impaired financial stability during the observed period, i.e., that they were in danger of bankruptcy. In addition, based on the results obtained using the Beneish M-score model, it was identified that a number of companies showed signals that indicate possible fraudulent financial reporting. Further, it was found that less than half of the observed companies reported on environmental protection in their annual reports, and they did so by providing a modest amount of information. The originality and value of the paper lies in suggesting that policymakers in the Serbian emerging markets should pay more attention to the operations of companies from the observed sector, as well as to their financial and non-financial reporting. Future research should focus on comparisons with agricultural companies from the same sector whose securities are listed on stock exchanges in the region
Mathematical Approach for System Repair Rate Analysis Used in Maintenance Decision Making
No full textReliability, the number of spare parts and repair time have a great impact on system availability. In this paper, we observed a repairable system comprised of several components. The aim was to determine the repair rate by emphasizing its stochastic nature. A model for the statistical analysis of the component repair rate in function of the desired level of availability is presented. Furthermore, based on the presented model, the approach for the calculation of probability density functions of maximal and minimal repair times for a system comprised of observed components was developed as an important measure that unambiguously defines the total annual repair time. The obtained generalized analytical expressions that can be used to predict the total repair time for an observed entity are the main contributions of the manuscript. The outputs of the model can be useful for making decisions in which time interval repair or replacement should be done to maintain the system and component availability. In addition to planning maintenance activities, the presented models could be used for service capacity planning and the dynamic forecasting of system characteristics
Eliminating Disparate Impact in MCDM: The case of TOPSIS
No full textIn today's business, decision-making is heavily dependent on algorithms. Algorithms may originate from operational research, machine learning, but also decision theory. Regardless of their origin, the decision-maker may create unwanted disparities regarding race, gender, or religion. These disparities may further lead to legal consequences. To mitigate unwanted consequences one must adjust either algorithms or decisions. In this paper, we adjust the popular decision-making method TOPSIS to produce utility scores without disparate impact. This is done is by introducing "fairness weight" that is used for the calculation of the utility function of TOPSIS method. Fairness weight should provide the smallest possible intervention needed for a decision without disparate impact. The effectiveness of the proposed solution is shown on the synthetic dataset, as well as on the exemplar dataset regarding criminal justice