This paper analyzes the constitutional position of Bosnia and Herzegovina (BiH) between constitutional nationalism and human rights. In its judgments in Sejdić and Finci v. BiH, and Zornić v. BiH, the European Court of Human Rights (ECtHR) found that the BiH Constitution was not in line with the European Convention on Human Rights and Fundamental Freedoms (ECHR). Namely, the privilege of the three constituent peoples in the Constitution of BiH, the ECHR was assessed as contrary to the prohibited discrimination prescribed by the ECHR. In this sense, the paper analyzes the current Bosnian constitutional model of a form of consociational democracy or constitutional nationalism. The authors analyze the presumption of the ECtHR in terms of the (non) existence of a legitimate goal for maintaining discriminatory provisions in the Constitution of BiH. Also, the paper analyzes the importance of the implementation of the ECtHR judgment for the future of the rule of law in BiH.

Concepts developed in the field of political economy offer a useful framework for explaining, from a western perspective, the phenomenon of Trumpism as an expression of anti-system politics emerging rich democracies in response to the effects of neoliberal growth models and the cartel forms of political parties. While insightful, this theory does not account for the role of media in presumed relationships. The article aims to offer, through exploratory analysis, the theoretical basis for conceptualization of the links between media, anti-system politics, and growth models. Trumpism developed, we argue, under the conditions of destabilization of traditional institutional arrangements of the news business, which enabled and limited the functions of journalism important for the democratization of the „Western“society. The paper contributes to the development of an alternative theoretical approach to the study of the media aspect of Trumpism concerning the prevailing corpus of ideas about "post-truth", "false news" and "echo chambers".

The most prominent concerns of contemporary British literature have been reserved for the revision of tradition and history and contestation of metanarratives through historiographic metafiction and historiographic metadrama. Liz Lochhead’s works are abundant in elements of historiographic metadrama which Lochhead uses to rewrite (hi)stories from a different angle, especially (hi)stories involving famous women and their position in the society, as is the case with Blood and Ice and Mary Queen of Scots Got Her Head Chopped Off. Blood and Ice focus on Mary Shelley’s process of writing her novel Frankenstein while Mary Queen of Scots got her Head Chopped Off presents Mary Queen of Scots and Elizabeth I in the light of their strained relations. Pertaining to Blood and Ice, the aim of this paper is to discuss the position of MaryShelley as a woman artist surrounded by Romanticists such as P.B. Shelley and Lord Byron and their liberal humanist ideology which shows great indebtedness to the patriarchal metanarrative. With regards to Mary Queen of Scots Got Her Head Chopped Off, the paper examines MaryStuart and Elizabeth I’s roles as women and monarchs, masculinity-femininity dichotomy surrounding the queens, the problematics of their historical representation, as well as the danger of their mythologization. The analysis of the elements of historiographic metadrama in the two plays shows that they are examples of ‘herstories’ that dismantle male-centered narratives as imposed rather than natural.

This study investigates interactive behaviors and communication cues of heavy goods vehicles (HGVs) and vulnerable road users (VRUs) such as pedestrians and cyclists as a means of informing the interactive capabilities of highly automated HGVs. Following a general framing of road traffic interaction, we conducted a systematic literature review of empirical HGV-VRU studies found through the databases Scopus, ScienceDirect and TRID. We extracted reports of interactive road user behaviors and communication cues from 19 eligible studies and categorized these into two groups: 1) the associated communication channel/mechanism (e.g., nonverbal behavior), and 2) the type of communication cue (implicit/explicit). We found the following interactive behaviors and communication cues: 1) vehicle-centric (e.g., HGV as a larger vehicle, adapting trajectory, position relative to the VRU, timing of acceleration to pass the VRU, displaying information via human-machine interface), 2) driver-centric (e.g., professional driver, present inside/outside the cabin, eye-gaze behavior), and 3) VRU-centric (e.g., racer cyclist, adapting trajectory, position relative to the HGV, proximity to other VRUs, eye-gaze behavior). These cues are predominantly based on road user trajectories and movements (i.e., kinesics/proxemics nonverbal behavior) forming implicit communication, which indicates that this is the primary mechanism for HGV-VRU interactions. However, there are also reports of more explicit cues such as cyclists waving to say thanks, the use of turning indicators, or new types of external human-machine interfaces (eHMI). Compared to corresponding scenarios with light vehicles, HGV-VRU interaction patterns are to a high extent formed by the HGV’s size, shape and weight. For example, this can cause VRUs to feel less safe, drivers to seek to avoid unnecessary decelerations and accelerations, or lead to strategic behaviors due to larger blind-spots. Based on these findings, it is likely that road user trajectories and kinematic behaviors will form the basis for communication also for highly automated HGV-VRU interaction. However, it might also be beneficial to use additional eHMI to compensate for the loss of more social driver-centric cues or to signal other types of information. While controlled experiments can be used to gather such initial insights, deeper understanding of highly automated HGV-VRU interactions will also require naturalistic studies.

Analysis of reports published by the leading national centers for monitoring wildfires and other emergencies revealed that the devastation caused by wildfires has increased by 2.96-fold when compared to a decade earlier. The reports show that the total number of wildfires is declining; however, their impact on the wildlife appears to be more devastating. In recent years, deep neural network models have demonstrated state-of-the-art accuracy on many computer vision tasks. In this paper, we describe the design and implementation of a lightweight wildfire image classification model (LW-FIRE) based on convolutional neural networks. We explore different ways of using the existing dataset to efficiently train a deep convolutional neural network. We also propose a new method for dataset transformation to increase the number of samples in the dataset and improve the accuracy and generalization of the deep learning model. Experimental results show that the proposed model outperforms the state-of-the-art methods, and is suitable for real-time classification of wildfire images.

During the 2020/2021 academic year, the working conditions of students during online classes, the levels of depressive, anxiety, and stress symptoms, and their correlations with resilience and sleep quality were examined on a sample of students from the University of Sarajevo (UNSA). The results of the research show that most students have satisfying working conditions from home, but also that a large number of them still feel fear and anxiety that something will happen to the internet connection or computer during online classes and exams. About 50% of students have increased symptoms of depression, anxiety, and stress, which are significantly in correlation with poorer sleep quality. Higher levels of resilience in students are correlated with lower levels of depression, anxiety, and stress and better sleep quality. The level of student depression is a variable that is the best predictor in explaining the criterion variable quality of sleep.

When lexemes are borrowed from a foreign language they go through different phases of integration, typically divided into four types: phonetic, orthographic, morphological, and semantic. The question of which gender to assign to a loanword and which gender assignment criteria to apply is still ambiguous in linguistics. Therefore, the aim of this paper is to determine the regularities of gender assignment to lexicalized and non-lexicalized anglicisms in German within linguistics and glottodidactics. In addition, the present study also investigates the question of whether deviations in gender assignment, in the case of lexicalized anglicisms from our corpus, can be explained by their semantic differences. The corpus for our research is composed of a total of 194 scientific articles, containing a wide range of linguistic topics and issues in the field of glottodidactics. The analysis showed that different criteria are used to assign different genders. Thus, in the case of feminine and neutral nouns, the suffixal analogy prevails, while in the case of masculine nouns, semantic analogy and monosyllabicity play a greater role.

Abstract: The paper presents results of the measurements of the sulfur dioxide (SO2) and nitrogen dioxide (NO2) concentration and meteorological parameters: temperature, air pressure, relative humidity and wind speed. The data were collected from January 2019 to December 2020 at two stations, namely Center and Heating plant, in the City of Bijeljina, Republic of Srpska, Bosnia and Herzegovina. SO2 and NO2 are one of the major air pollutants that could negatively affect the human health. Levels of SO2 and NO2 in air samples and meteorological variables from urban zone of Bijeljina were determined at both localities, which represent a highly-populated area with intensive traffic. This topic has not been studied up to now in Bijeljina, although the recent research data indicates that there is a correlation between meteorological parameters and air pollutants. Statistical analysis confirms direct corelation between SO2 and NO2 and meteorological parameters, specially temperature in locality Center (r = -0.639), the wind speed in locality Heating plant (r = 0.399) and relative humidity (r = 0.162). Correlation of NO2 with temperature is not confirmed in both localities. The wind speed increase is followed by rises of the NO2 concentration values and vice versa. Correlation of NO2 with pressure is confirmed in locality Center (r = 0.128) but it is not confirmed in locality Heating plant. Correlation between NO2 and relative humidity found to be negative in locality Center (r = -0.062). These parameters are the most important meteorological factors influencing the variation in SO2 and NO2 concentration in the air during the research. Depending on the obtained correlation, meteorological parameters had a positive or negative impact on air pollution.

The aim of the study was to determine the differences in the level of fat, sugar and body structure based on the level of functional abilities. The sample of respondents are students of the University of Bihać who also completed the shuttle run test (BEEPT). T-test for independent samples revealed the existence of statistically significant differences between the arithmetic means of the two groups of subjects for (AMAS, p = 0 .011), (BMI p = 0 .000), (FAT%, p = 0.000), (FMKG p) = 0.000). A statistically significant difference was also found in the variable triglycerides (TRIGL p = 0.019), while in the other variables no statistically significant difference was found in the two groups of subjects in favor of subjects who had better results in functional abilities. The coefficient of discriminant canonical correlation is (0.512), as is Wilks lambda, (0.738), which indicates very high discrimination between groups (sig. 000). The greatest contribution to the formation of the discriminatory function was given by the variables FMKG - .801, FATPR -.760, BMI - .707, AMAS - .390, TRIGL - .358, HOLE - .235. The centroids of the groups show a large distance between the results of the groups because they are located at both ends of the coordinate system. The first group consists of positive results of a total of 7 variables, which means that the respondents of the first group had significantly better results in these variables. Based on the results, it can be concluded that the increase in cholesterol triglycerides and some parameters of body structure affected the level of health status as well as body composition in students.

Artificial intelligence technology is rapidly advancing year by year, and thus the possibilities of using artificial intelligence in every branch of science. The importance of artificial intelligence and its components has been known for a long time. They are seen as tools and techniques that make the world a better place With their simple and everyday techniques, they make the world a mistake-free place. These technologies and applications are not only related to our general and everyday life, but also affect and have significance for other domains as well. The paper will analyze the application of artificial intelligence and its influence today.

In this paper, we deal with the potential applications of Artificial Intelligence in educational institutions. One of the main goals of applying the mentioned technology is to make the education system more efficient if possible. Based on research in this area in the world, our impression is that there is a long-term plan for introducing AI in schools, as a new kind of learning standard. The idea is not for artificial intelligence to completely take over the domain of education, but for some processes to be better organized or improved, so that the result of learning would be of better quality. New technologies, in the form of modern learning platforms that incorporate AI as one of the most important components of the system, significantly improve the communication and research work of students. It is evident that new generations are coming whose experience will be significantly different from ours, and that the new technologies from the domain of AI that we consider in this paper will significantly affect the design processes in the future.

Estimating conditional average treatment effects (CATE) is challenging, especially when treatment information is missing. Although this is a widespread problem in practice, CATE estimation with missing treatments has received little attention. In this paper, we analyze CATE estimation in the setting with missing treatments where unique challenges arise in the form of covariate shifts. We identify two covariate shifts in our setting: (i) a covariate shift between the treated and control population; and (ii) a covariate shift between the observed and missing treatment population. We first theoretically show the effect of these covariate shifts by deriving a generalization bound for estimating CATE in our setting with missing treatments. Then, motivated by our bound, we develop the missing treatment representation network (MTRNet), a novel CATE estimation algorithm that learns a balanced representation of covariates using domain adaptation. By using balanced representations, MTRNet provides more reliable CATE estimates in the covariate domains where the data are not fully observed. In various experiments with semi-synthetic and real-world data, we show that our algorithm improves over the state-of-the-art by a substantial margin.