ABSTRACT. While the evolution of digital technologies in human-related aspects changes the approach to organisational issues, artificial intelligence enables complex decision-making and supports strategically important evaluations. Management behaviour, decisions and activities at all organisational levels cause consequences of varying degrees. Recent developments across management related processes require a paradigm shift regarding the application of assistive technologies. Hitherto, the trends in evaluation of leadership parameters have been rather unpredictable and restricted to the application of a limited number of technologies. Leadership as a phenomenon is multifarious. In our exploration, we limit our scope of investigation to the degree of leadership as one of the decisive components for successful entrepreneurship and ranks as one of the organisational development indicators. In this paper, we discuss the current situation exhibited in publications on researched topic and propose a holistic approach to predict the defined leadership parameters over time, based on a regression decision tree model. In order to evaluate our proposed approach, we present selected implementation examples pursuing the identified goals of analysis. Subsequently, we discuss the proposed approach with a focus on the potential benefits, ob-stacles, limitations and perspectives.

ABSTRACT. This paper investigates whether the global trend of ownership concentration of international fi-nancial institutions can also be observed for the Deutschland AG, which is the informal designa-tion for the historically grown and largely isolated network of German stock listed companies. Us-ing network analysis, capital linkages of German HDAX and SDAX companies in both 2006 and 2018 are analysed and the results are compared. The network analysis enables a systematic presen-tation of the capital linkages and also helps to analyse link strengths and make supposedly hidden relationships visible. The results show a noticeable increase in the concentration of internationally active investment companies in the ownership structures with a simultaneous decline in the par-ticipation rates of German investors in German companies. Therefore, both the trend of owner-ship concentration of international financial institutions and the erosion of the Deutschland AG can be confirmed.

ABSTRACT. The article contains a description of the process and results of the implementation of the k-means algorithm in the analytical platform Loginom for the problem of clustering the regions of Ukraine by the level of investment attractiveness in the field of tourism. The selection of tourism clusters and their ranking is a difficult task in the field of data analysis, as there is no single consolidated indicator of investment attractiveness. The conclusion about the affiliation of a particular region to one of the tourist clusters is determined by a set of indicators of the volume of tourist services for different types of economic activity in the field of tourism. The Loginom system has powerful tools for cluster analysis using EM-Clustering, k-means, g-means and others. The tools of statistical and visual analysis of the obtained results deserve special attention: Table, Statistics, Chart, OLAP-Cube, Cluster Profiles. Clustering has made it possible to identify groups of regions that are actively developing the tourism industry (primarily Kyiv city and Odesa region) and are currently formed for tourism investors. Equally important is the selection of problem regions that have a low level of attractiveness for domestic and foreign tourism. It is noted that Ukraine has a huge potential for the development of the tourism industry. The regions that, according to the results of the cluster analysis, are in the problem group, in fact, have "world-class tourist pearls". The Government of Ukraine and local authorities should pay attention to the insufficient level of development of the tourism industry, provide comprehensive support to the regions that are in the problem cluster, and thus increase their level of investment attractiveness.

ABSTRACT. This paper presents a program system called an algebraic virtual machine (AVM), which handles industrial hardware specifications, programs in different languages, and models in algebraic language. It uses the formal algebraic methods that were developed in the scope of behavior algebra and help to resolve the problems of verification, analysis, testing, and cyber-security. It permits the possibility of creating your own methods and theories and trying them with industrial examples with minimal efforts. The machine learning technique is used for the definition of formal method efficiency, and the classification model is trained during algebraic processing. The formalization and checking for resistance of blockchain attack is considered.

ABSTRACT. Like the human brain, an artificial neural network is a complex nonlinear parallel processor for processing information; it is often called a neurocomputer. Classical systems of artificial intelligence are always naturally associated with classical logic and discrete mathematics. Thus, the representations and models of knowledge, undeniable at least since Aristotle, do not correspond to the cognitive models that are obtained as a result of studying the human brain. In view of Niels Bohr, quantization is a phenomenon of a discrete, sequential process, that inherent in continuous and stochastic systems. However, the traditional mathematical model of quantum mechanics did not imply generalization to dissipative systems. The corresponding generalization, called the Dynamic quantum model (DQM), was proposed by author. It is defined for any dynamic system, given by ordinary differential equation or diffeomorphism or for dynamic systems that using logical operations. The neural network is exactly the DQM in the space of input signals. In this paper DQM is defined and constructed universally for both Hamiltonian systems and systems with the fuzzy logic truth function on phase space. The paper goal is to demonstrate quantization on DQM, i.e. actually on neural networks, and to extend the classical Bohr-Sommerfeld condition to the general case, in particular, to systems with a fuzzy truth function.

ABSTRACT. The authors present a new approach to conflict management to ensure proper collaboration between different aviation personnel using decision making methods in uncertainty. To improve the results of collective decisions, dual risk assessment of decision making in an emergency is used. Initially, the operators' decision is influenced by the factors of occurrence and development of an emergency. The next collective matrix is formed from the individual rational decisions of the operators. The reliability and optimality of the result solutions are provided by the individuals and collaborative solutions of operators. The optimal solutions in emergency “Failure of one engine on a twin-engine aircraft” using collaborative-factor decision making models for the pilot, flight dispatcher, and air traffic controller are obtained.

ABSTRACT. As we know, the least attention is paid to pregnant women in our state. Therefore, many of them have to absorb all the information they come across, which is not reliable, and has not been veried. As a result, these women nd themselves in dicult and dangerous situations, because they try and experience all sorts of recommendations or act in their own way, which leads to dierent outcomes: they can no longer get pregnant, infant mortality, have a negative impact on their own or on the health of the baby. Therefore, it is important to provide them with informational assistance and support during pregnancy. The scope of the mobile application is focused on patients of medical organizations in the healthcare sector. The application is aimed at fast interaction and high-quality communication with the operators of medical centers. And also it provides reliable information for the population of the Republic of Kazakhstan in accordance with all the requirements and regulations. The "Healthy Mom" mobile application on Android OS is intended for all pregnant women (and who are planning) in the Republic of Kazakhstan.

ABSTRACT. Reliability-based unmanned aerial vehicle (UAV) fleet nuclear power plant (NPP) monitoring mission planning models with battery recharging are developed. Battery recharging is carried out either at the depot or by using autonomous battery maintenance stations deployed at certain points. A classification in accordance with ways for UAVs to follow their routes and recharge their batteries is given. An example of the proposed models application is given. The probability of the successful fulfillment of the plan for the UAV fleet to perform the NPP monitoring mission is used as an indicator when using the proposed models.

ABSTRACT. One of the hardest problems in sentiment analysis is an analysis of multiple targets in the same text, without clear boundaries between target contexts. This problem is even harder for Ukrainian and Russian languages because of the lack of datasets and established approaches. Responding to the business needs of our company, we created the bilingual dataset, manually annotated for targeted sentiment according to strict guidelines. This dataset allowed us to fine-tune a pre-trained multilingual BERT model and to achieve large improvement in key metrics (macro F1, F1 for negative and positive classes) over the baseline models. As a by-product, we have trained new NER models for both languages. NER and targeted sentiment models were successfully introduced into the production environment at Semantrum.

ABSTRACT. The paper presents the results of the comparison of two nonparametric methods of authorship identification of the Ukrainian literature texts. The paper describes the implementation of the corresponding methods based on the Klyushin–Petunin tests and its simplified version. The method of n-gram selection is applied. For testing a collection of texts up to 200,000 characters from 10 authors was used. As a result of carrying out the test, it was found out that the simplified test appears to be more sensitive and specific, and monograms and bigrams in opposite to trigrams provide clear detection of authorship.

ABSTRACT. We introduce a new lexicon of discourse connectives for the Ukrainian language. Discourse connectives like ‘because’, ‘therefore’ are grammatical elements which link clauses and sentences semantically and play a crucial role in discourse structure. They have shown to be useful for many tasks in natural language processing from argumentation mining to authorship analysis.

We introduce a semi-automatic method for inventorizing discourse connectives in underresourced languages, by leveraging existing lexicons from other languages. As a result, we provide the first computer-readable lexicon of 129 Ukrainian discourse connectives. We provide syntactic as well as semantic information for these items. Finally, we carry out a small pilot study using the lexicon for discourse level corpus annotation, and report on the distribution of connectives in Ukrainian in two different types of media.

ABSTRACT. This paper considers business process model structuredness issues, which are mostly related to inaccurate usage of gateways. According to related work in the process model structuredness domain, split gateways ought to match respective join gateways of the same type, while the existing mismatch measure allows evaluating model structuredness only by degrees of split and join gateways. Thus, the current measure of process model structuredness is not accurate enough and process model shortcomings may remain undetected, which may affect negatively model understandability, maintainability, and increase the error probability of business process models. Hence, error fixing costs may grow exponentially during later stages of the information system lifecycle. Therefore, we have proposed an improved gateway mismatch measure and a model to detect design issues and suggest changes necessary to achieve a sufficient level of business process model structuredness. The software tool for business process model structuredness analysis was developed to perform experiments with a large set of business process models of different industries. Analysis of obtained results, including sample business process models, detected design issues, and estimated efforts and cost-saving benefits are outlined. Conclusions were made, and future work was formulated.

ABSTRACT. The volume of private investment is growing steadily nowadays. In this case, it is crucial to analyze investors' behaviour, decision-making factors and the specifics of their investment portfolio formation and especially their cognitive constraints, which prevent them from effectively defining investment goals and profitable achieving them. This research shows that investors, even experienced and financially literate, often make significant mistakes when creating their own investment portfolios. Thus, the use of automated tools for determining the insurance premium and the optimal investment portfolio, which is a robo-adviser, becomes relevant. The paper presents the model for estimating personal insurance premium for different risk attitude investment portfolios using robo-advisor. Three types of investors are analyzed: con-servative, aggressive, and moderately aggressive. The model helps determine the individual size of the insurance premium for each investor profile, taking into account his or her risk attitude.

ABSTRACT. International commodity, financial and foreign exchange markets operate under conditions of uncertainty, incomplete or asymmetric information. Consequently, their development is nonlinear and does not fir fundamental laws, and therefore cannot be sufficiently predictable. In addition, the export of goods depends on different factors that form multidimensional data sets. It requires sophisticated analytical tools such as Data Mining and Machine Learning to conduct analysis and make accurate management decisions.

ABSTRACT. Paper describes the creation of software for automated assessment of pro-jects’ investment attractiveness. Research goal is development of software for automated assessment of investment attractiveness of projects using in-vestment criteria. Research methods include algorithms for assessing the in-vestment attractiveness of projects using defined criteria, software develop-ment on the BAS ERP platform using finance criteria. To integrate these cri-teria scoring system for selection of investment projects was implemented. Comparative analysis of the software for assessing the investment effective-ness of projects was carried out, the main criteria of assessing the investment attractiveness of projects were determined, and a software tool for automated analysis of the investment attractiveness of projects was developed to evalu-ate different projects taking into account definite criteria.

ABSTRACT. We propose a unified induction proof principle which is applicable to situations that arise in analysis and verification of nondeterministic discrete, continuous, and hybrid systems. This induction principle generalizes Noetherian and real induction.

ABSTRACT. The approach of algebraic modeling of molecular interactions in some environment to determine the triggering of the studied properties is considered in the present article. The main idea of this study is to represent the actions of elementary particles in different molecular structures, in particular the motion of electrons in orbitals as algebraic equations for further processing. Behavior algebra specifications are used as the modeling language. The article also describes the formalization of the examples of atoms interaction (creating of chemical bonds) on the example of Ionic Bond. The formalization and properties analysis is considered with the usage of the insertion modeling platform. The study is at an early stage of development and the approach is demonstrated by some examples.

ABSTRACT. The paper is devoted to the problem of checkability of circuits in FPGA components of safety-related systems, which are designed to operate in two modes: normal and emergency for providing their own functional safety and the safety of control facilities in order to prevent accidents and reduce losses in case of their occurrence. Functional safety is ensured through the use of fault-tolerant solutions that are sensitive to sources of multiple failures, including hidden faults. They can accumulate during a prolonged normal mode with limited checkability of the circuits and simultaneously manifest themselves with the beginning of the emergency mode. A fault-tolerant structure becomes fail-safe if it is checkable. The problem of hidden faults manifests itself in the memory of the LUT units of FPGA components with LUT-oriented architecture. The program code written in the memory of the LUT units is checked with a checksum, but it can be corrupted when reading its bits on the outputs of the LUT units. Bits observed only in emergency mode reduce the checkability of FPGA components and are potentially hazardous. Checkability can be increased by the operation of circuits on successively replaced versions of the program code that can be obtained for the same hardware implementation. Versions move potentially hazardous bits to checkable positions observed in normal mode. However, the set of these versions are significantly limited by connecting the inputs of the LUT units to the inputs of the FPGA component. The proposed method overcomes this limitation by introducing an additional scheme. Experimental studies of library FPGA designs show a low level of their checkability and efficiency of the proposed method, which provides totally checkable circuits.

ABSTRACT. Today the problem of botnets detection is very actual, as botnet are widespread and are used to perform different types of cyberattacks and to cause threats to network services and users' properties. One of the means the botnets use to connect with their command-and-control (C&C) is the domain name system (DNS). On other hand, the fust-flux technique enables to avoid botnets' detection. The paper presents a new botnets' detection technique, which takes into account, the DNS feature analysis, botnets' architecture aspects, as well as their behaviour in the network and hosts. The proposed approach allows detecting the botnets' bots of centralized, decentralized and hybrid architecture with high efficiency.

ABSTRACT. Modern information and communication technologies (ICT) are vulnerable to APT-attacks (advanced persistent threats) and other relevant threats. APT-attack is a stealthy threat actor, typically a nation-state or state-sponsored group, which gains unauthorized access to ICT and remains undetected for an extended period. Early detection of such threats is a relevant and important task. In this paper, a method of linguistic terms using statistical data was used for structural and analytical models of parameters (both host and network parameters). Based on this, logical rules can be developed to provide the functioning of IDS based on honeypot technology for APT-attacks detection and intruder type identification in ICT.

ABSTRACT. The current state of the problem regarding secure control of smart mobile objects using existing telecommunication infrastructures is described in the paper. The scientific and technical task of developing the principles of using cellular commu-nication systems for the organization of secure remote control of intelligent mo-bile objects has been solved. The solution lies in the use of overlay computer networks based on VPN tunneling. Proposals have been developed for construct-ing an overlay computer network based on VPN taking into account traffic ag-gregation as well as dividing network elements according to their purposes. The related problems of connection dependability at the access, distribution and core levels are considered. The possibility of using nested VPN tunneling in high-speed cellular communication systems to improve the security of transmitted data is analyzed. The result of a number of model and experimental studies is the proof of the proposed principles efficiency. Recommendations have been devel-oped for using the proposed principles of constricting an overlay network based on cellular communication systems for the secure control of intelligent mobile ob-jects in solutions related to the implementation of dependability and resilience of computer networks concept.

ABSTRACT. Cardamom is an IRC-funded project related to the development of natural language processing (NLP) tools for minority and historical languages. At its core, the project builds on the comparative principle, which entails you can learn information about a language by looking at either its close sibling languages or, in the case of historical languages, its modern form. For this goal, we are focussing on building the largest collection of known resources on languages in these families by searching the web, social media and retro-digitization. Then we are creating new deep learning models which will be applied to under-resourced languages to generate new dictionaries and computer-aided language learning tools, and to historical languages to provide a workbench to analyse historical language data. We will describe progress towards these goals and the recent collaboration with Translators without Borders to develop language resources for the Rohingya refugee crisis.

ABSTRACT. In this paper, the different models for the estimation of both local and global coherence of Ukrainian-language texts have been considered. In order to evaluate the local coherence of a document, Transformer-based and LSTM neural networks have been proposed with further training on a Ukrainian-language news corpus. It has been shown that the LSTM-based approach outperforms the corresponding network based on the Transformer architecture according to the accuracy metrics while solving typical tasks on both test datasets. In order to investigate the connection between sentences revealed by the neural network, the Uniform Manifold Approximation and Projection dimension reduction technique has been utilized for the projection of sentences’ embedding into 2D space. The clusters obtained may indicate the consideration of both the structure of a sentence and different types of connections between them by the designed model. In order to estimate the global coherence of a document, a model based on a graph convolutional neural network has been suggested. The appropriateness of taking into account the connection between all sentences despite their positions has been shown. The results obtained for the designed and trained global coherence estimation model may indicate the different aspects of the analysis of a text by the designed models that can lead to the usage of both local and global coherence estimation models according to an assigned task.

ABSTRACT. One of the crucial problems of natural language processing for languages such as Ukrainian is lack of labeled datasets both for pretraining of word embeddings or large deep learning models and for benchmarking existing approaches.

In this paper we describe a framework for simple classification dataset creation with minimal labeling effort. We create a dataset for Ukrainian news classification and compare several pretrained models for Ukrainian language in different training settings.

We show that ukr-RoBERTa, ukr-ELECTRA and XLM-R large tend to show the highest performance, although XLM-R large tends to perform better on longer texts, while ukr-RoBERTa performs substantially better on shorter sequences.

We publish this dataset on Kaggle (https://www.kaggle.com/c/ukrainian-news-classification/) and suggest to use it for further comparison of approaches for Ukrainian text classification.

ABSTRACT. In this paper, at the first time, we perform the analysis of correlational and non-extensive properties of the CO2 emission market relying on the carbon emissions futures time series for the period 04.07.2008-10.05.2021 and the daily data of the power sector from the U.S. Carbon Monitor for the period 01.01.2019-10.05.2021, which consist the data of both individual countries (USA, Germany, China, India, United Kingdom, et al.) and global emissions (World). For comparison, we present the analysis of the Dow Jones Industrial Average (DJIA) index. Our results show that both futures and the DJIA are presented to be non-extensive, and the distribution of their normalized returns can be better described by power-law probability distributions, particularly, by q-Gaussian. We estimate Tsallis triplet for the entire time series of CO2 emissions futures and the DJIA and present q-triplet as an indicator of crisis phenomena, relying on the sliding window algorithm. It can be seen that the triplet behaves in a characteristic way during economic crises. The toolkit of random matrix theory (RMT) allowed us to investigate the correlational nature of the carbon emissions market and to build appropriate indicators of crisis phenomena, which clearly reflect the collective dynamics of the entire research base during events of this kind.

ABSTRACT. Research goals and objectives is to design an algorithm for identification of factors influencing the price dynamics of shares at the time of publication of quarterly reports. Subject of research: design forecast algorithms for trade platforms at stock exchanges. Research methods used: comparative analysis, business analytics, design of software module, correlation regression analysis, analytical methods. Results of the research: The paper contributes to the empirical studies of in-terim companies’ disclosure in the equity prices fluctuation and to practical design of algorithms for trade forecasting platforms. The key finding is that the highest price dynamics was observed during the quarterly reports. The largest impact on price movements during the quarterly reports was given by the number of open short positions and capitalization of companies.

ABSTRACT. We researched how to use financial technology in the finance industry on the example of robo-advisors; defined the basic functionality of robo-advisor; got the robo-advisors implementation based on analysis of the most popular financial services. We compared their functionality, composed a list of critical features and described the high-level architectural design of a general robo-advisor tool, scope of application of robo-advisors, their key features, and a brief overview of existing solutions. Using Markowitz model, we set up a concept of a robo-advisor application for investors who have differ-ent attitudes towards risks. Our goal is to cover the main features of financial robo-advisor and to describe a high-level architecture for such applications. We have defined the main modules that represent the architecture of a typical robo-advisor. We also describes different techniques, which could be applied for building a personalized investment and insurance portfolio.

ABSTRACT. The effectiveness of the introduction of interactive teaching methods, such as simulation games in ERP-systems, in the educational process appeared to be promising both for traditional and online learning, thus, the world expertise should be learned and disseminated in order to increase the competitiveness of Ukrainian education. The aim of this article is to develop models for applying simulation games in ERP-systems in order to exploit multidisciplinary education-al opportunities in economic educational programs and to enable students of Ukrainian HEIs to master professional competencies both in face-to face and online learning. A specific feature of this study is to demonstrate the models for implementing the approach to the study of the courses related to quantitative methods of decision-making support. We suggested to analyze the compliance matrix of the program competencies and the components of the educational program, to identify mandatory and elective courses that develop statistical and analytical skills, to find correspondence between processes and tasks embedded in games, and methodological support of decision-making, to identify relevant topics and conduct classes using simulation games, which will allow students to apply the acquired knowledge in real situations to make informed managerial decisions. The implementation of the proposed idea is presented in terms of Economic Cybernetics educational program. Two models of ERPsim implementation in the educational process are proposed, the advantages of each model are analyzed. Examples of the use of data from ERP systems for the implementation of quantitative methods of decision-making support in such courses as Operations Research in Economics and Methods of Decision Justification in Economics un-der Information Uncertainty are given.

ABSTRACT. The research objective is to is to develop an algorithm to computerise the process of allocating the limited resources of a local government to maximise the needs of the community. With limited financial resources, local govern-ments must determine the optimum volume of planned services to be pro-vided. The increasing amount of information, as well as the need for rapid management decision-making, necessitates the use of information and com-puter technologies (ICT) in this area. Research methods used: comparative analysis, planning theory, utility analy-sis, design of software module, analytical methods. Results of the research: The paper contributes to the theoretical studies about ICT implementation in local governance. Also, the paper contributes to the discussion of the practical implementation of ICT in the allocation of limited resources at the local governance level.

ABSTRACT. Advertising is a unique socio-cultural phenomenon: its formation is due to social, psychological, linguistic factors, features of the “aesthetic consciousness” of society and its cultural traditions. Advertising text – a special kind of text, it is a carrier and expression of information. It is important that the text is interesting to the desired audience, and it can be formed by methods of text semantic analysis and highlight the keywords on the basis of which advertising content can be generated. In this regard, it is developed an intelligent method of advertising content forming of higher education institutions on the semantic analysis basis and thus advertising manager can generate advertising content. The implementation of the method was carried out on the basis of a survey of “Computer Science” students regarding admission. Semantic analysis of documents based on LSA and LDA-method is performed. The results show that more than six keywords are present in document 0, based on the LSA method – 66%. Based on the LDA method, the vast majority of keywords are presented in document 2 – 82%. Based on the obtained keywords, the LSA and LDA methods created content for advertising of higher education institutions. The effectiveness of the generated advertising content on the basis of LSA and LDA-method was compared, a comparative experiment was conducted on Facebook on the business page “Computer Science of ZUNU”. By effectiveness comparing results of generated advertising content, the effectiveness of the ad increased by 44% and the price for the result decreased by 31%.

ABSTRACT. Accelerated technological development in the context of the Fourth Industrial Revolution changed the nature of competition in world markets, increasing the importance of technological opportunities as a source of competitive advantage and identifying technology as a key factor in production. Every year, digital technologies change everyday life, creating the foundations for sustainable socio-economic development. Changes resulting from the revolution in information technology signal the need for new approaches to training, particularly in Ukraine. Technologies are improving at a fairly rapid pace, but the me-thodological base at the level of Ukrainian high education institutions (HEIs) is adapting to such changes rather slowly, which, accordingly, slows down the process of "smarting" of education. In turn, graduates are not the most attractive for the modern labor market. This article highlights the urgent need for extensive training in this area, and, in turn, offers a case of the study programme for graduating data science analysts (DSAs). The original approach is that the case is for the social science faculty but not for engineering faculty. The necessity of DSAs is extremely high in the economic field / business however mostly gra-duates of the engineering faculties having strong programming skills lack the economic knowledge and understanding of business laws. The proposed pro-gram differs from existing ones on the market, but not implemented in HEIs, with its systematical adaptability to the requirements of the state standard, as well it meets all the requirements of employers in the field of Data Science.