This study provides a comprehensive statistical analysis of the uptake and application of artificial intelligence (AI) technologies in the European Union (EU) member states up to 2023. The research draws on data from 150,400 companies to examine the relationship between different AI technologies such as machine learning, process automation and text mining. Using correlation, factor and principal component analysis, the study explores the extent and effectiveness of the integration of technologies, providing a new scientific perspective on the industrial application and strategy of AI-based innovations. The analysis has revealed that countries with higher levels of digital skills and advanced technological infrastructure, such as Denmark and Finland, exhibit significantly higher AI adoption rates. Furthermore, the results highlight how closely certain technologies, such as machine learning and robotic process automation, are related. The results offer significant contributions to facilitate a more effective application of AI technologies in the European industrial environment and provide guidance for future development strategies.

Only a few sections of the Hungarian Academy of Sciences (HAS) maintain ranked journal lists, one of which is the IX Section of Economics and Law. The doctoral committees within this section assess candidates for the title Doctor of the Academy of Sciences based on eight lists of ranked international journals. These lists remain unchanged for a fixed period (three to five years). This study examines the extent to which the average SCImago Subject Ranking (SJR) values differ across subject categories within the economic subject areas of the SCImago database To analyze these differences, we apply various models of analysis of variance (ANOVA), including Welch’s ANOVA and the Kruskal-Wallis ANOVA.

This study focuses on the problem of stock management, a part of the circular economy, considering reuse. The basis of our presented model is the well-known Leontief model, where we investigate the possibility and impact of collection and reuse processes. It also follows that the concept of a circular economy can be approached as a problem or task in several ways. Using the Leontief model's concept, we examine the dynamic case. The new result of this study is that we can show the effect of distinguishing between newly produced and reused goods on final consumption in the reuse process. We show that stock dependence is relevant to closing the circular economy model. The analysis of the discrete dynamical system shows the growth potential of reuse in the production process. In the model, the dynamization solves the waste management problem: the collected but unused goods appear as stock in the next production period. The originality stems from the dynamic, stock-based, and dual-sub-economy extension of the Leontief model, specifically tailored to address reuse, stock management, and waste minimization in the circular economy, filling a recognized gap in the existing literature. We illustrate the results with numerical examples.

The ecological footprint has been a crucial ecological indicator for more than two decades, and the methodology for calculating it has developed significantly over the years. However, some issues and shortcomings still need to be addressed and specified further. This paper focuses on the embedded land requirements of imported commodities in input-output modelling approaches. We propose a refined model to overcome the shortcomings of two former models. Our model quantifies the embedded ecological land-footprint of imported commodities and their allocation between direct final consumption and production. In addition, we allocate the latter again among final consumption and exports using the framework of linear algebra and matrix arithmetic. We also propose ways of extending the model to overcome the general but misleading assumption in the literature that imported commodities have an equal per unit ecological footprint to domestic products, an approach that is based on the idea that trading partners have the same technological background.

This paper introduces a firm-level digital maturity index for small and medium enterprises (SME-DMI), created using an entropy-based objective weighting method. The indicator was developed through a representative survey of Hungarian firms and its main objective is to evaluate the digital applications, tools and skills used by companies. It encompasses digital tools and infrastructure access, digital application usage and related skill levels. In addition to the main dimensions and their associated weights, this indicator also analyses the relationships between digital dimensions and firm size, using ANOVA. The results demonstrate that the impact of firm size is significant for both digital skills and knowledge and for business applications. However, connectivity and access digital public services is no longer a distinguishing factor between small and large firms.

Our paper is based on the five principal dimensions of the International Digital Economy and Society Index (I-DESI), but instead of using the original scoring model based on arbitrary pre-determined weights, we apply more objective ranking methods that use the statistical properties of the data series to determine where the Visegrad Group (V4) countries (Czechia, Hungary, Poland and Slovakia) stand in terms of digital development among the countries of the European Union and other developed countries in the data set. The ranking is performed using the DEA-CWA (Data Envelopment Analysis/Common Weights Analysis) method (with six models) and the TOPSIS (Technique for Order of Preference by Similarity to Ideal Solution) method. Although the resulting weight vectors differ significantly from the arbitrary weights set by the European Commission, the country rankings remain similar, displaying relatively little sensitivity to the weighting method chosen.

The pursuit of sustainable development has emerged as a pivotal concern in contemporary business discourse. Sustainable development is increasingly important for businesses, including micro-enterprises, which must address sustainability challenges to stay competitive and viable. A key aspect of corporate sustainability is understanding stakeholder dynamics and translating their needs into business practices. Our research, using a questionnaire, examined the relationship between sustainability pressures from six key stakeholders on Hungarian microenterprises. We also explored whether distinct groups of variables could be identified based on the perceived directions of these sustainability pressures. Our findings, grounded on optimal scaling and multidimensional scaling analyses, reveal a clear link between how micro-enterprises monitor stakeholder activities and the focal points of these stakeholders' impact concerning sustainability. Managers must prioritize stakeholders based on their influence, whether through legal regulations, social values, or financial factors. This prioritization helps micro-enterprises effectively target their sustainability efforts, enhancing performance and long-term success.

The European Union's climate neutrality ambitions pose significant challenges for member states in terms of energy production, consumption and greenhouse gas emissions. The 2030 targets set out packages of measures that provide both a justification and a timely opportunity to explore the relationship between economic performance and certain indicators of energy management, with a view to answering the question of how the relationship between the different variables has evolved over time: have we seen a structural change or is the change more gradual? To this end, the present study examines the relationship between the relevant indicators (GDP, energy production/use, greenhouse gas emissions) of the 27 EU member states using multivariate statistical methods for the year 2012 and compares it with data from an analysis already carried out for the year 2022 (Er & odblac;ss et al. 2025). In addition to presenting the changes identified using descriptive statistical tools, the study will use correlation analysis to assess the relationship between the individual indicators and principal component analysis to identify the key factors that play a role in the evolution of each variable. Finally, the authors use cluster analysis to highlight the changes that have taken place over the 10 years, by showing the differences between the country groupings. The results of the research do not show significant structural changes in the relationships between the variables, but both descriptive statistical tests and correlation analysis yield favourable figures for energy consumption and GHG emissions, which are below those registered for growth in economic performance. The results of the cluster analysis from 2012 to 2022 show a strengthening of the convergence of the member states based on the variables analysed. 2025

Economic growth, which is the focus of nations' central objectives, is causing significant environmental costs and damage (Chou et al., 2023). Fossil energy use contributes to economic growth but is also a major source of carbon emissions and accelerating rates of climate change, requiring governments to balance spending on economic growth with sustainable energy management (Bhuiyan et al., 2022). This research aims to examine and evaluate the links between economic performance and energy management. This will be achieved through a multivariate statistical analysis of relevant EU data (GDP, energy production and consumption, energy exports and imports, GHG emissions) for 27 Member States in 2022. The results of the correlation analysis will partly provide insights into the relationships between the variables, while the results of the principal component analysis will allow for identification of background variables. The cluster analysis shows a high degree of homogeneity among members, but several outliers can be identified.

AbstractThe paper employs a cross-sectional data set comprising the main dimensions of the European Union's International Digital Economy and Society Index (I-DESI) and utilises grouping methods based on objective weights to evaluate the relative digital readiness of Hungary and other Central and Eastern European (CEE) member states of the EU. The objective was not to establish a total ordering (ranking) of the countries in the data set, but rather to identify the most appropriate means of grouping the CEE countries into homogeneous units, utilising multivariate statistical and decision-theoretical techniques (tiered DEA, partially ordered sets and clustering). Despite the disparate methodologies employed, the findings are consistent in that the CEE countries (including Hungary) exhibit a general resemblance to one another and demonstrate comparatively lower levels of digital readiness than Northern and Western European countries. The notable exception is Estonia, which exhibits a distinctive level of digital advancement.

While EU countries are acting as one and making commitments to achieve global climate targets, there are significant differences in the performance of individual countries and thus their contribution to the targets. The present study aims to establish a ranking based on objective weighting, using relevant environmental and economic indicators, with the main objective of identifying the position of the V4 countries (Czech Republic, Poland, Hungary, Slovakia). The ranking is carried out using the DEA (Data Envelopment Analysis) methodology with 5 different approaches (stepwise envelopment analysis, three different models of the common weights approach, stepwise Pareto efficiency approach). Despite the different weighting schemes of the different models, the country positions are well defined, on the basis of which the Member States can be grouped into five groups. The V4 countries tend to show signs of the least efficient structures, but the weight vectors in each model allow the reasons for this to be identified.

Tanulmányunk célja az Európai Unió digitális gazdaság és társadalom fejlettségét mérő mutatója (Digital Economy and Society Index, DESI) 2020. évi kiadásának öt fő dimenziója, illetve két makrogazdasági mutató (AIC [tényleges egyéni fogyasztás] és GDP/fő [bruttó hazai termék]) összekapcsolt elemzése többváltozós statisztikai módszerekkel. Vizsgálataink két részre oszthatók. Elsőként a változók közötti lineáris összefüggéseket vizsgáljuk egyszerű Pearson- és kanonikus korrelációs elemzés, regressziós elemzés, valamint faktoranalízis révén. Itt elsősorban arra a kérdésre keressük a választ, hogy a digitális átalakulás előrehaladását, annak dimenzióit milyen módon és mértékben befolyásolja a gazdasági fejlettség. Ezután az Európai Unió tagállamait a digitális-gazdasági fejlettségük alapján csoportosítjuk klaszterelemzés és többdimenziós skálázás révén. Ennek eredményei szerint az uniós országok három jól elkülönülő csoportra oszthatóak, amelyek közül az elsőbe a gyengébben fejlett, főként kelet-közép- és dél-európai periféria, a másodikba a fejlett nyugati-északi államok, a harmadikba pedig két éllovas tartozik.

The selection of input and output items is crucial for successful application of Data Envelopment Analysis (DEA) as they should express the decision maker's preferences and perceptions of what might affect the efficiency of a decision making unit (DMU). This article addresses the question of the transformation of input and output data that may be required for efficiency analyses using DEA method. Different methods for the data transformation are available in the literature, however, they may lead to different results, which may bias the decisions. This paper attempts to provide some guidance on this issue and to compare the results. An example of supplier evaluation will be used to illustrate the possible solutions and the differences in the final results (supplier evaluated to be among the efficient suppliers).

The rankings of universities around the world were created with the aim of measuring the performance of higher education institutions as well as the quality of institutions. Such lists provide a basis for better informed decisions by applicants to the higher education market but can also be an important source of information for decision-makers in individual states on how each country's institutions are performing in the field of international scientific competitiveness. There are numerous examples of such rankings. The authors aimed to select a leading one, namely, the QS World University Rankings 2021, and to examine the ranked universities based on statistical variables obtained from the Scopus and SciVal databases created by the Dutch academic publishing company Elsevier. Thus, the aim was not to rank institutions but instead to closely examine the statistical variables and criteria of the universities ranked by QS with the help of multivariate statistics. The results show that Scopus/ SciVal data can be used to examine not only researchers but also universities. The results also show a high degree of similarity.

The objective of the paper is to develop an analytical tool that is capable of modelling decision-making in coopetitive business relationships. Managers in the same industry differ in respect of their willingness to adopt coopetition. To better understand coopetitive decision-making, we need a model whereby such decisions can be experimented with and analysed. An important prerequisite of such a model would be its capacity to measure the performance consequences of coopetitive interactions at both firm and relationship levels. We show that existing operationalisation has limited capacity to do that. Based on existing game theoretical constructs, we propose a new operationalisation of a coopetitive decision-making episode in horizontal business relationships using a two-step sequential game. We suggest developing what we term a “coopetitive composite solution matrix” by summing up the payoff functions of the two steps of the game. The suggested operationalisation has the capacity to measure all the potential performance consequences of a complex piece of coopetitive decision-making in an episode. In this way, the decision problem’s cognitive representation becomes straightforward and analysis of the impact of the behavioural attributes of managers on the actual decision-making process is unambiguous.