At the present stage, the achievement of the set strategic goals of ensuring Russia's economic independence and technological leadership is associated with the development and implementation of domestic information and cognitive technologies. The country's agro-industrial complex, which is undergoing a complex process of digital transformation with the expansion of the use of robotic technology and intelligent systems, plays a special role in solving the basic tasks of maintaining state sovereignty. The development of platform solutions in agricultural production faces serious limitations and constraints on the effective application of the "digital twin" concept, due to unresolved issues regarding the conceptual and institutional justification for their construction for organizational systems. In this regard, the aim of this study is to substantiate proposals for defining the concept of a digital model of an agricultural enterprise and the formation of a possible option for describing the economic system and basic business processes for conducting full-cycle smart agriculture. The application of content and logical analysis methods, and reengineering technology, allowed us to appropriately define a reference digital model of an enterprise in the agricultural sector and present a possible design for a digital model of the economic system of a smart agricultural enterprise. Definitions of the concepts of "digital model" and "digital twin" for organizational systems are proposed, clarifying existing definitions in terms of reflecting the variability of the description of the organization's business model when displaying the entities of "business architecture" and "business processes" as separate structural elements and the contour of subjective perception of information when making decisions. The structure of a digital model of an agricultural enterprise's economic system in a networked precision farming environment is substantiated, taking into account changes in the composition and role of production factors in a data economy. We demonstrate the need to reflect in this model elements and relationships that address the requirements of ensuring environmental neutrality and social responsibility in full-cycle agricultural production. We recommend using the information image of a digital twin of an agricultural enterprise to design the structure and fill the model of the economic system with data based on regulated forms of planning and reporting documentation when building a digital platform to support management decision-making. The digital twin ontology description scheme expands our understanding of the theoretical foundations of the methodology and tools for designing and developing information models of objects and processes for business systems.
The rapid development of artificial intelligence (AI) is accompanied by increasing computational complexity and decreasing model transparency, which significantly limits its adoption in critical domains that require a high level of trust, interpretability, and justification of decisions. Under these conditions, the field of Explainable Artificial Intelligence (XAI) has gained particular importance as it focuses on approaches and technologies that enable understanding of AI system logic and interpretation of their outputs. This article examines the timely topic of implementing XAI in the context of Industry 5.0. Special attention is given to practical application scenarios: the authors present concrete industrial cases from IBM, Siemens, and other companies demonstrating how XAI contributes to enhancing the reliability, safety, efficiency, and trustworthiness of AI systems. The study includes a systematic search and analysis of the literature in this domain and proposes well-grounded key criteria for comparing existing XAI approaches. The article also outlines the advantages, current limitations, and promising directions for the development of XAI, highlighting the opportunities it opens for improving effectiveness, transparency, and trust in business.
Amid intensifying competition in the B2B e-commerce sector, particularly within the Do-It-Yourself (DIY) segment, traditional static search architectures increasingly suffer from limited adaptability and declining retrieval relevance. This study examines the limitations of rule-based ranking approaches and proposes a dynamic product ranking framework based on the Learning-to-Rank paradigm implemented with LightGBM. The primary objective of the research is to quantitatively evaluate the economic return on investment (ROI) associated with the deployment of personalized ranking algorithms. A simulation-based digital experiment was conducted using a synthetic user clickstream model to approximate real-world interaction behavior. The results indicate that the proposed dynamic ranking model yields significant improvements in search effectiveness, as measured by the metric, while simultaneously generating quantifiable gains in key business performance indicators. Specifically, the implementation resulted in a 2.1 percentage point increase in the conversion rate and a 14.5% uplift in incremental revenue. These observed effects achieved statistical significance. These findings provide empirical evidence supporting the economic viability of transitioning from static search systems to intelligent ranking architectures, highlighting their strategic importance for scalable and competitive B2B e-commerce platforms.
This paper presents a description of the current state and the results of an analysis of recent advances in the problem domain of automated task distribution among employees. The purpose of the study is to identify the main trends and patterns in the development of existing task allocation methods, to determine their strengths and limitations, and to justify the need for new approaches and algorithms that can improve the efficiency of task delegation to employees. Using a unified system of notations for the key concepts of the subject area, the article provides a concise descriptive review of ten universal task distribution algorithms published over the past twenty years. The comparative analysis was carried out according to a set of criteria reflecting both the technical and the organizational-behavioral aspects of how these algorithms function. The key evaluation criteria included: the degree to which performer competencies are taken into account; adaptability to changing external conditions and team composition; requirements for completeness and structure of the input data; robustness to incomplete or noisy data; transparency and explainability of decision-making; computational complexity; scalability with an increasing number of tasks and employees; implementation and maintenance costs; and orientation toward personnel development and competence enhancement. The comparative analysis we carried out made it possible to identify the advantages and shortcomings of each method and to formulate recommendations for their most effective practical application. The results showed that none of the examined algorithms can be considered a universal tool for delegation. Furthermore, it was found that comprehensive information about a performer's suitability for solving tasks requiring diverse competencies is either ignored or insufficiently utilized by many algorithms. This observation leaves open the problem of developing new approaches to task allocation and designing new algorithms based on them.
Modern challenges in organizational security, particularly within critical infrastructure sectors (energy, transportation, finance, IT), necessitate innovative solutions to mitigate risks associated with hiring unreliable personnel. This requires a shift from conducting fragmented checks to the creation and implementation of comprehensive systems for proactive risk assessment. The urgency of developing such systems is driven by the high frequency and catastrophic consequences of insider incidents, coupled with the inability of traditional methods to detect complex, multi-stage threats originating from employees. However, building intelligent systems that semantically integrate heterogeneous data (biographical, behavioral, financial, digital) presents new systemic challenges. The aim of this article is to analyze the key methodological, ethical-legal, and architectural requirements for designing such systems. The work sequentially examines: 1) ethical and legal dilemmas (fairness, privacy, the right to explanation) and the constraints imposed by personal data legislation; 2) specific cyber threats targeting the compromise of the knowledge base and system logic, along with architectural countermeasures based on Security by Design principles; 3) a comparative analysis of the technological components of a multi-level assessment system (documentary verification, psychometric testing, AI analysis), justifying the necessity for their integration. The scientific novelty lies in a synthetic approach that forms a holistic methodology, considering not only technological efficiency but also fundamental legal constraints and information security requirements. The practical significance of the work consists in formulating systemic requirements for the design of secure, lawful, and socially responsible intelligent decision support systems for personnel security.
In the context of the increasing need to improve the management efficiency of enterprises that support the implementation of the principles of digital transformation based on the concept of the fifth-generation industry, the relevance of research on the development of appropriate systems in terms of ensuring continuous targeted and sustainable development, customer-centricity and social orientation of production is increasing. Digital twin technology and its multi-agent implementation act as effective means of building enterprise performance management systems. At the same time, the lack of scientific research in this area determines the purpose of the article, which is to develop a product-resource approach to enterprise performance management based on digital twins in the fifth-generation industry. A distinctive feature of the proposed approach developed by the authors is the use of dynamic enterprise performance management technology based on digital twins, which ensures the integration of business processes and resources used at the level of not only one enterprise, but also at the level of network value chains based on a common digital platform of the business ecosystem. The paper analyzes approaches to the intellectualization of enterprise management, on the basis of which the requirements for an enterprise performance management system are formulated, ensuring the solution of interrelated tasks of targeted enterprise development, the formation of flexible value chains, and the rational and sustainable use of enterprise resources. The possibilities and disadvantages of the efficiency management process in EPC class systems are analyzed. The paper substantiates the use of digital twin technology and its multi-agent implementation to build an enterprise performance management system in the context of mass customization and the network nature of value chains in the fifth-generation industry. A process for managing the efficiency of enterprises at all stages of the life cycle based on the technology of digital twins of products and resources has been developed, dynamically ensuring the targeting, adaptability and sustainability of the functioning and development of the enterprise.
The Russian Business Intelligence (BI) technology market has undergone significant transformations under the influence of import substitution policies and geopolitical shifts. The departure of major foreign vendors such as Microsoft Power BI, Tableau, and Qlik has reshaped the landscape, increasing the demand for domestic BI solutions. This study examines the current state of the Russian BI market, analyzing the challenges and opportunities associated with transitioning to local platforms. While global trends emphasize cloud adoption, self-service BI, and AI-driven analytics, the Russian market faces specific barriers, including regulatory constraints on cloud services, a shortage of skilled specialists and the limited functionality of domestic solutions compared to their international counterparts. The research employs a comparative analysis of global and domestic BI platforms, assessing their advantages and constraints in the context of import substitution. Key Russian BI solutions such as Foresight Analytical Platform, Visiology and Yandex DataLens have demonstrated potential in filling the gap left by foreign vendors, offering functionalities tailored to local business needs. However, challenges remain in areas such as usability, scalability and integration with enterprise IT infrastructure. The study also explores the role of government initiatives and corporate investments in accelerating the development of competitive domestic BI technologies. Findings indicate that despite the difficulties of adaptation, the Russian BI market is evolving through increased digitalization, growing enterprise demand for data-driven decision-making and the necessity of developing independent analytics ecosystems. Addressing issues such as improving user training, expanding platform capabilities and fostering collaboration between IT developers and businesses will be crucial for the successful advancement of BI technologies in Russia. This study contributes to the ongoing discourse on the evolution of the Russian BI market, emphasizing its role in supporting business competitiveness and economic resilience in a rapidly changing technological and regulatory environment.
Predicting the financial insolvency of firms is crucial for investors, creditors and regulators. However, access to high-quality, balanced data for model training is often limited due to privacy concerns, information scarcity or financial reporting characteristics. This paper explores the potential of synthetic data generation techniques to increase minority class instances in unbalanced datasets and thereby potentially improve insolvency prediction models. The paper compares the performance of various imbalance reduction methods, including established methods such as, for example, the Synthetic Minority Oversampling Technique (SMOTE), with new synthetic data generation approaches based on Bayesian networks, marginal distributions, random forests and generative adversarial networks. The performance of these methods is investigated in terms of their ability to improve classification performance such as Gini coefficient, geometric mean, false positive and false negative rate. The sample for the experiment is real financial performance of industrial SME companies in Finland for 2021. The results contribute to the growing body of knowledge on synthetic data generation and its application to address imbalanced datasets and improve predictive modeling in the financial industry and provide insights into the effectiveness of different synthetic data generation methods for sampling imbalanced datasets and improving the accuracy and reliability of firm insolvency prediction models.
This paper considers the transformation of methods and models of strategic management based on the ecosystem approach within the framework of the formation of a unified digital platform for management. The ecosystem approach to the socio-economic development of society is gaining popularity as a result of global social requirements for environmental protection and a careful attitude to use of limited natural resources. Environmental problems in the Russian agricultural sector are increasing, in particular due to the process of formation of agro-industrial associations, mainly in the form of agricultural holdings. In this case, there is a problem of a systematic approach to using of technologies for the integration of all types of resources involved in production, taking into account the growing number and importance of environmental factors. Mathematical modeling is proposed as the main method of strategic management research. Unlike most of the existing models, which are often iconographic, the model proposed allows us to consider a larger number of factors. This makes it possible to assess various options of the modeling objects development using a simulation approach. As a result of the research, a mathematical model of strategic management of agroholdings for sustainable development was developed. It is shown that the development strategy should be implemented considering an appropriate automated management information system. This will lead to a radical change in the whole system of management and production. It will allow the enterprise to apply strategic goal-setting, focusing first of all on quality, controllability and other components of competitiveness. The mathematical model proposed provides justification of unified methods of long-term digitalization applicable for large agricultural associations, as well as for small and medium-sized farms, which will be able to cooperate with agricultural holdings relying on the principles of outsourcing.
In this paper we analyze the problem of intelligent product matching in digital marketplaces for which one requires evaluation of similarity of various records that describe products but may differ in format, content or volume of multimodal data. The subject area of this scientific research represents an intersection of entity resolution (ER) problem solving methods: record matching and multimodal data analysis. It is of extreme relevance in a fast-growing platform economy with the e-commerce market expanding exponentially. The main purpose of this research is to develop and test an intelligent multimodal model based on transformer architecture to improve the accuracy and robustness of product matching in digital marketplaces. The authors developed a model integrating textual, visual and tabular attributes which enables us to identify similar products, find competitive offers, detect duplicates and perform product clustering and segmentation in a more effective manner. The proposed approach is based on the self-attention mechanism which enables contextual-semantic relations modeling of various-nature data. In order to extract the vector representation of text descriptions, language models are applied, in particular the Sentence-BERT architecture; for the graphical component Vision Transformer is used; and tabular data are processed using specialized learning mechanisms based on TabTransformer structured data. The experiment we carried out demonstrated that the developed multimodal model efficiently solves the task of product matching in digital marketplaces in an environment of significant variability of product items and data heterogeneity. Additionally, the results suggest that the model can be adapted successfully for application in other product categories. The results obtained confirm the efficiency and expediency to apply the multimodal approach for digital marketplace product matching implementation. This allows the e-commerce market participants to significantly improve the quality of inventory management, increase pricing efficiency and strengthen their competitive advantages.
Megaprojects represent large-scale investment programs with complex organizational structures, uniting a multitude of stakeholders whose interactions lead to the redistribution of power and the creation of temporary management centers. In conditions of unstable and uncertain external environments, such stakeholder behavior can result in the failure to achieve the set goals of the megaproject. An important scientific task is the development of mathematical models and methods for managing changes in megaprojects caused by the integrative actions of stakeholders under complex external conditions. The present study is aimed at creating a mathematical model and developing an information system for neural network analysis of the intensity of changes in megaprojects. Megaproject management is described using a vector-matrix model of a dynamic system with feedback based on the results of changes. To identify recurring patterns of negative events, the event-oriented analysis method was used. This allows for justifying new approaches to management aimed at reducing uncertainty and enhancing the effectiveness of megaproject implementation. Based on the proposed tools, a retrospective neural network analysis of the intensity of changes in the "Nord Stream 2" megaproject was conducted. Within the study, key groups of stakeholders were identified whose interactions significantly impacted the project's implementation: Group 1-Gazprom PJSC, European companies and the governments of Russia and Germany supporting the project; Group 2-the governments of transit countries, the USA, environmental organizations and Baltic region countries opposing the project or expressing concern about its consequences. It was demonstrated that the integration of separate stakeholder groups contributes to the formation of temporary management centers with varying interests, leading to an increase in both positive and negative changes within the project. The outcome of the work was the development of an information system for analyzing the intensity of changes in megaprojects in the form of a prototype, which includes: a mathematical model for managing changes in megaprojects; a neural network analysis methodology based on the use of a large language model for processing textual information and generating quantitative assessments; as well as a software interface for uploading documents, automated data processing, and visualization of results. The primary neural network used was the large language model Qwen 2.5-Plus, which, while not specifically adapted for this task, had its parameters calibrated for analyzing the intensity of changes in megaprojects. The system prototype provides users with the ability to analyze stakeholder interactions, assess the intensity of changes and forecast potential risks based on historical data. A promising direction for further research involves applying the model we developed and neural network analysis methodology for comparative studies of various types of megaprojects.
The increasing pace of development of e-commerce continues to present new challenges in terms of personalizing product search and recommendations. Monolithic search and recommendation systems have become cumbersome and are unable to effectively address the need for a deeper understanding of users on electronic trading platforms (ETPs) despite having access to comprehensive information about their interests and purchase histories. Collaborative filtering mechanisms which are widely used suffer from a lack of diversity in offerings and a reduced capacity to surprise users. Additionally, the low frequency of recommendation updates and the replacement of "personalized" with "similar to others" concepts contribute to these issues. We have approached the resolution of these issues by developing a shopping assistant named "Ellochka" that is individual for each user of ETP. The digital avatar model of the user continually searches for relevant products based on their history of interaction with ETP. We were guided by the principle of independence-avatar models do not share information with each other. When a new user joins, they are assigned a unique avatar model that evolves independently. Each avatar has its own language to generate search queries. The level of complexity of each avatar can vary depending on the intensity of its interaction with ETP. Continued interaction with the avatar allows for tracking of optimal purchase conditions, reminding users of expiration dates and the need for repurchasing frequently purchased items. Isolating the avatar allows it to be retrained after each event, without significantly impacting the overall search and recommendation system. The use of neural network architecture-based and Kolmogorov-Arnold networks in the avatar-model has led to improvements in the main indicators of search and recommendation effectiveness, namely, novelty and diversity.
This article considers ways to improve the efficiency of the regulated procurement market by implementing recommender systems into the existing procurement IT infrastructure. Using state, municipal and commercial procurement of electric power products as an example, the article considers promising classes of recommender systems for implementation, proposes a methodology for developing such services, and discloses algorithms for processing, configuring and interpreting data necessary for their operation. The difference between the author's approach to creating services and previously published works is substantiated, testing and A/B testing are carried out, and an assessment of the effectiveness is presented. The results obtained have scientific novelty (the methodology of using neural networks in relation to the procurement industry has been substantiated) and practical significance (the customer's time saved on searching for suppliers by up to 40%; the pool of potential suppliers has been expanded; supplier risks have been diversified by selecting relevant procedures from new areas and from new customers; suppliers have been provided with the opportunity to find up to 2-3 new customers for 1 recommendation mailing with a frequency of 1-2 times a week). We proposed to implement the developments in the practice of the operator of public procurement tenders. The authors see further development of recommendation services and solutions for the procurement industry in improving the analysis of semantic (text, logical) content of procurement documents, as well as the behavioral strategies of suppliers. The risks and limitations are associated with the high cost of maintaining a staff of developers-practitioners in neural networks, possible hallucinations of neural networks and their
The simplex method is widely used in economic planning and forecasting tasks. However, this method is used in real economic activity to find solutions to large-scale tasks, the speed of which is not a critical factor. This significantly limits the applied value of the simplex method in the economic sphere, since currently there is a certain tendency to move to more detailed economic models, which makes it urgent to accelerate calculations based on the simplex method. In these conditions, GPU (Graphical Processor Unit) computing accelerators become the most important means of accelerating calculations. The authors propose the implementation of the simplex method in matrix form for computing on GPUs using the PyTorch library. This allows you to switch to using the computing power of graphics processors in a simple and reliable way. A linear programming problem with 900 constraints is solved on a graphics accelerator 6-9 times faster than the solution on a conventional processor. This paper identifies groups of applied economic problems for which the proposed algorithms and methods can be relevant.
In mass real estate valuation, in cadastral valuation, there is a problem of splitting the value of a single real estate object into the value of land plot and buildings (improvements) located on it. One of the key information sources for real estate valuation is market data. Such data may contain information on offer prices, as well as actual transaction prices (for example, in mortgage transactions) for the whole object. At the same time, in the accounting policy of enterprises different rates of land and property tax often require separate accounting of the value of land plots and the buildings located on them. The problem of such splitting of a single object’s value is the subject of permanent discussions in the valuation community. There are no established methods. This article proposes a method of splitting the value of a single property object based on the approach borrowed from co-operative game theory. A simple game formulation of the problem and its fair solution based on the Shepley value are considered. Simple and well-interpretable computational formulas are obtained, which allow us to split the market value of single objects on large data sets in minimum time. The proposed method is new in the theory and practice of valuation.
This article is devoted to the problem of determining the sequence of project implementation in a program of improving business processes of an organization. The relevance of the study is related to current conditions, where the quality of business processes is essential not just for the success, but also for the survival of an organization. Improvement of business processes is a costly program that involves certain projects. The projects of the program cannot be started at the same time due to limited budget and human resources. Thus, we face the task of determining the sequence of stages of program implementations. The solution of this task is one of the most important problems of business informatics. This paper proposes a new criterion for prioritizing projects. It takes into account the fact that funds for projects are generated during the implementation of business processes. The criterion also takes into account the pace of spending the project budget and the need for participation of key employees of the organization. The implementation of the program is divided into a few stages. At each stage, the problem is solved by determining a set of projects whose sum of priorities is maximum and whose resource requirements do not exceed the constraints developed at that stage. The relevance of the article is initiated by looking at the need of enterprises that ensure the airworthiness of civil aviation airplanes. This work is of interest for project program managers of production and service companies, as well as for a wide range of researchers.
The processes of formation and development of intellectual capital in the digital economy are ill structured processes occurring in conditions of a significant increase in the speed and unpredictability of changes in the external environment. This makes it extremely difficult to use previous experience and probabilistic forecasts when assessing the risks of failure to achieve strategic goals for the development of the intellectual capital of an organization. At the same time, undesirable deviations in achieving these goals can lead to significant negative consequences. In this regard, there is a need to develop appropriate fuzzy methods and models, all of which determines the relevance of this work. The purpose of this study was to develop a fuzzy method for assessing the risks of failure to achieve the strategic goals of an organization in the field of intellectual capital development. The method is based on a fuzzy model developed by the authors which allows us to take into account the uncertainty tolerance of the decision maker. Testing the method on the example of a specific organization showed the possibility of its practical applicability. We provide quantitative assessments and qualitative interpretations of the risk levels of failure to achieve target indicators for the development of the intellectual capital of an organization (a large regional university).
Modern cities are facing increasing traffic congestion, necessitating the implementation of intelligent traffic management systems. One of the key areas in this field is adaptive traffic signal control, which can adjust to changing traffic conditions. However, existing methods for optimizing traffic signal cycle parameters have several limitations, such as high computational complexity, the risk of premature convergence of algorithms and the difficulty of accounting for traffic dynamics. This study proposes an approach to optimizing the characteristics of an intelligent transportation system using hybrid evolutionary algorithms. The methods we developed combine the principles of genetic algorithms (GA) and particle swarm optimization (PSO), enabling a balance between global and local search for optimal parameters. The research examines six different hybridization schemes, including modified versions of basic algorithms, as well as their integration with HDBSCAN clustering methods for adaptive optimization frequency tuning. To evaluate the effectiveness of the proposed algorithms, a simulation model was developed in the AnyLogic environment, replicating real urban traffic conditions. Numerical experiments conducted on a local section of the road network in Moscow demonstrated that the hybrid SlipToBest algorithm achieves the best results in reducing average travel time and fuel consumption, while the Alternating algorithm ensures high solution stability. The results of this study confirm the feasibility of using hybrid evolutionary methods for traffic flow management tasks. The proposed algorithms not only enhance the efficiency of traffic signal control but also establish a foundation for the further development of adaptive urban traffic management systems.
Recommendation systems are widely used in the commercial field. The algorithms and architectures of recommendation systems are similar in various fields of application and have proven their effectiveness. Recommendations are based on the user’s profile, the manner of his behavior on various IT (Information Technology) resources, as well as on similar users. At the same time, the use of recommendation systems in specialized areas is not widespread. Technology divisions are a promising new area of application for recommendation systems, and IT experts themselves will be the users. The purpose of this article is to consider a combination of a recommendation system, machine learning (ML) and LLM (Large Language Model) and to design these tools in a single system. Data volumes are currently measured in petabytes (1015 bytes) and exabytes (1018 bytes). In order to process even technical information (metadata/technodata) from the surrounding IT landscape, from the IT systems used by experts, AI (Artificial Intelligence) agents are needed. This article provides a literature review regarding the use of recommendation systems in combination with LLM applications, and suggests an application architecture model that generates human-readable news from technical event logs. The system is designed for a group of users who work with big data (ML engineers, data analysts, and data researchers). It is a combination of recommendation system technologies, LLM, and machine learning models. The article also provides the first results of the research that was carried out.
This study was conducted within the framework of the urgent task of studying the processes of developing the human capital of an organization and increasing employee productivity. At the same time, the development process is viewed through the prism of creating and implementing various elements of the well-being program into the main corporate business processes of the organization. The purpose of this work is to develop a fuzzy method for forming an optimal portfolio of well-being program activities which will allow you to get as close as possible to the target values of key performance indicators (KPIs) of employees on a given planning horizon. To achieve this goal, a hypothesis is put forward about the possibility of building a tool that allows, based on the functional dependencies of influence channels, to form an optimal portfolio of well-being program activities that increases the efficiency of the organization. The method developed consists of a model representing a fuzzy programming problem and a method for finding its solution. A distinctive feature of the model is the consideration of two levels of uncertainty in the formation of an optimal portfolio of activities related to the reliability of estimates of numerical coefficients of functional dependencies of channels of influence and a set of parameters of constraints determined by experts. An integral indicator is used as the target function of the model, which characterizes the degree to which the target values of key employee performance indicators are achieved, taking into account the importance of each of them for the organization. The optimization variables in the model are binary variables that determine the inclusion of a certain event in the well-being program of an organization at a specific time within a given planning period. The limitations in the model are: the total amount of financial resources allocated for the implementation of the well-being program; the amount of investment in a specific area of the well-being program; an increase in the integral indicator of competence of each employee. From a practical point of view, the proposed method will make it possible to form a well-founded portfolio of well-being program activities, the implementation of which has the maximum possible positive impact on employee productivity.
