CC BY
31
Данные об авторе
Сидорова Антонина Васильевна,
д.э.н., профессор, Донецкий национальный университет им. Василия Стуса, завкафедрой бизнес-статистики и экономической кибернетики e-mail: 100 prima@gmail.com
Data about the author Antonina Sydorova,
Doctor of Economics, prof. Donetsk National University named after Vasily Stus, Head of the Department of Business Statistics and Economic Cybernetics e-mail: 100 prima@gmail.com
UDK 311.213 http://doi.org/10.5281/zenodo.3829617
БУРК1НА Н.В.
Вплив використання великих даних на украшський 6i3Hec
Управл'1ння корпоративною ¡нформац1ею мае на уваз1 П оперативний зб'р, ефективну органза-цю та оптимальне використання на благо б\знесу. Поряд 3i структурованими даними, як можна з'брати i впорядкувати в таблицях, велика частина ¡нформацИ' представлена величезною ктьюстю документв, електронних листв, в'щео та ¡ншим неструктурованим контентом. В1н не менш важ-ливий, але мае на уваз\ бльш детальн методи систематизацИ та застосування нових методов i технологй обробки даних. Один з них - Big Data. За рахунок використання великих даних компанИ' можуть отримувати вщчутн конкурента переваги. Сучасний обсяг даних та 'х збльшення зростае з кожною секундою, що обумовлюе актуальнсть досл'щжень в сфер1 застосування Big Data i ме-тод\в анал'зу великих даних для прийняття оптимальних управлнських р1шень для б'знесу. Метою даноI статт'1 е систематизаця як класичних, так i сучасних методов анал'зу даних, як можна засто-сувати в р'зних економ'нних сферах з метою полпшення показникв ведення б\знесу. У статт'1 були поставленi i виршен наступн завдання: пошук рол'1 i мсця великих даних для б 'знесу; виявлення методологИ застосування науки Data science до ведення укра'нського б\знесу в Укран система-тизаця методов анал'зу даних i з'ясування особливостей Iх адаптацИ' до укра'нського б\знесу.
В робот запропонована система методов аналтики сучасних даних, що включае як класичн методи досл'щження структурованоI ¡нформацИ', так i сучасн технологи, як дозволяють вир'1шувати про-блеми б\знесу по-новому - дешевше, швидше або ефективнше. Запропонована система методов може бути застосована для анал'зу даних р'зних вид1в п'щприемств i органзацй при вир'1шенн'1 таких завдань як прогнозування ринковоI ситуацИ; маркетинг i оптим'защя продаж1в; утримання кл1ент1в, прогнозування динам'1ки цн вдосконалення продукцИ'; прийняття управлнських ршень; п'щвищення продуктивност'1 працк ефективна лопстика; монторинг стану основних фонд1в i багатьох ¡нших.
Ключовi слова. Big Data, Data science, анал'пика, методи анал'зу даних, б'знес-анал'пика, штучний ¡нтелект, машинне навчання.
БУРКИНА Н.В.
Влияние использования больших данных на украинский бизнес
Управление корпоративной информацией подразумевает ее оперативный сбор, эффективную организацию и оптимальное использование на благо бизнеса. Наряду со структурированными данными, которые можно собрать и упорядочить в таблицах, большая часть информации представлена огромным количеством документов, электронных писем, видео и прочим неструктурированным контентом. Он не менее важен, но подразумевает более детальные методы систематизации и применение новых методов и технологий обработки данных. Один из них — Big Data. За счет использования больших данных компании могут получать ощутимые конкурентные преимущества. На сегодняшний день объем данных и их увеличение растет с каждой секундой, что обуславливает актуальность исследований в сфере применения Big Data и методов исследования больших данных для принятия оптимальных управленческих решений для бизнеса. Целью данной статьи является систематизация как класических, так и современных методов анализа данных, применимых в разных экономических сферах с целью улучшения показателей ведения бизнеса. В статье были поставлены и решены следующие задачи: поиск роли и места больших
данных для бизнеса; выявление методологии применения науки Data science к ведению украинского бизнеса в Украине; систематизация методов анализа данных и выявление особенностей их адаптации к украинскому бизнесу.
В работе предложена система методов аналитики современных данных, включающая как классические методы исследования структурированной информации, так и современные технологии, которые позволяют решать проблемы бизнеса по-новому - дешевле, быстрее или эффективнее. Предложенная система методов может быть применена для анализа данных разных видов предприятий и организаций при решении таких задач как прогнозирование рыночной ситуации; маркетинг и оптимизация продаж; удержание клиентов, прогнозирование динамики цен совершенствование продукции; принятие управленческих решений; повышение производительности труда; эффективная логистика; мониторинг состояния основных фондов и многих других.
Ключевые слова. Big Data, Data science, аналитика, методы анализа данных, бизнес-аналитика, искусственный интеллект, машинное обучение.
BURKINA N.V.
Influence of Big Data and Business Analytics use on Ukrainian business
Corporate information management implies its prompt collection, efficient organization and optimal use for the benefit of the business. Along with quantitative, or structured, data that can be collected and organized in tables, most of the information is represented by a huge number of documents, emails, videos, and other unstructured content. It is no less important but implies more detailed methods of systematization and the application of new methods, technologies and data processing tools. One of them is Big Data. Due to big data, companies can gain tangible competitive advantages. Today, the volume of data and its increase is growing every second, which determines the relevance of research in the application of Big Data and modern methods of researching big data for optimal management business decision making. The purpose of this article is to systematize both classical and modern data analysis methods applicable in various economic fields in order to improve business performance. The following tasks were solved in the article: identifying the role and place of big data for business, include for Ukrainian entrepreneurship; identification of the methodology for the application of Data Science to conduct Ukrainian business in Ukraine; systematization of data analysis methods and identification of their adaptation features to Ukrainian business.
The paper proposes a system of methods for analyzing modern data, which includes both classical methods for studying structured information and modern technologies that allow solving business problems in a new way - cheaper, faster or more efficient. The proposed system of methods can be used to analyze data of different types of enterprises and organizations in solving problems such as forecasting the market situation; marketing and sales optimization; customer retention; forecasting price dynamics; product improvement; managerial decision making; increase in labor productivity; efficient logistics; monitoring the status of fixed assets and many others.
Keywords. Big Data, Data science, analytics, data analysis methods, business analytics, artificial intelligence, machine learning.
Every year, the amount of information generated by a person increases. According to analytical reports the volume of stored data will increase to 40-44 zettabytes by 2020 and it will increase up to about 400 zettabytes by 2025. As the result, the management of structured and unstructured data using modern technology is an area that is becoming increasingly important. In the process of the information boom and the development of new methods for
processing human-generated data, the term Big Data arose. Historically it is believed that the first time it was proposed in 2008 by the editor of Nature magazine - Clifford Lynch. Since, the Big Data market has grown by several tens of percent annually. And this trend, according to experts, will continue. Thus, according to Frost & Sullivan, in 2021, the total global market for big data analytics will increase to $ 67.2 billion. The annual growth will be about 35.9% [1].
During the massive spread of technology, people have generated a huge amount of data they are not able to process and visualize. Data on our calls and movements, behavior on the Internet, preferences in stores, people-made changes in the landscape, climate processes and many other things. They are all Big Data which can benefit greatly with proper processing. Big data analytics can identify extremely valuable information from structured or unstructured data sets. Thanks to it, a business, for example, can determine trends, forecast production indicators and optimize its own expenses. And in order to reduce costs, companies are ready to introduce the most innovative solutions [2].
Big data analytics is used by more than 50% of companies worldwide now. Even though in 2015 this figure was only 17%. Big Data is most actively used by companies that work in the field of telecommunications and financial services as well as companies that specialize in healthcare technology. Minimal use of Big Data analytics in educational companies: however, the most of this field representatives announced intention to use technology soon.
In the USA, Big Data analytics is used actively: more than 55% of companies from various fields work with this technology. In Europe and Asia, the demand for big data analytics is not much lower. It consists about 53%.
Big data is not just data. This is either a lot of data vertically (a large sample that is difficult to process with standard analytical tools) or horizontally (a large array of factors for which classical methods and models are generally not applicable).
For effective work with big data, another toolbox is needed, which it became machine learning in the era of the development of information technology. According to the methods of machine learning, a person only gives the computer some input, determines the way the machine is trained, but the machine learns itself; it comes to one or another answer and analyzes the information. Machine learning is not only artificial intelligence. This sphere includes genetic and evolutionary algorithms, simpler tasks related to cluster analysis [3].
The culture of businesses that were formed before the widespread of Big Data, Machine Learning and Data Science is usually based on a conservative approach to risk, minimizing costs, and increasing operational stability. Previously, companies tried to use only time-tested business models and tech-
nologies, and key competencies were often outsourced for the sake of economy. This was partly because the markets were mature. For example, banks knew well who their competitors were, what their place in the ranking was, what they would do in the next some years. The situation was more stable and predictable.
With the development of digital technologies, this situation has changed. Young and energetic companies have begun to appear that take away the market share. To survive in new conditions, it's necessary to scale thinking and arsenal of skills at all organization levels.
• At the top management level - to recognize the risks that inaction and the rejection of new technologies entail.
• At the level of middle managers - to develop general technological literacy, understanding of a digital product: what can be done with it, what cannot, what role does Data Science play in the business.
• At the level of data scientists - to develop professional skills to be able to quickly solve a difficult task as well as Real-time decision-making.
Real-time decision-making is providing information in the context and integration with the workflow in real time to any device and anywhere. So that decisions can be made at a particular time. Real-time decision-making plays a role not only in making decisions within the organization, but also in how to communicate with customers and understand their needs at any given moment. For example, a customer is in a shopping center and makes purchases. At this moment, he receives an offer to go to the store from the list of the program «Discounts for you», where he can get a good discount on the card. This offer is relevant to the client here and now, and not tomorrow, when he has already left the shopping center. Also tracked are purchases of air or train tickets.
The boundaries of what is possible when solving such problems are determined, in fact, by several factors: the creativity of the employees and the ideas they create, the quality and availability of the data necessary to bring the idea to life, and the adequacy of the technologies used in the company.
Data science = Data + Science
Data. The first component of data science, without which the entire further process is impossible. Data component consists of the following processes: how to collect, store and process data, as well as
how to extract useful information from a common data array. It is up to 80% of their working time to clear the data and bring it to the right kind. An important part of this issue is how to handle data for which standard methods of storage and processing are not suitable because of their huge volume and / or variety. They are big data. At the same time, data analysis specialists do not always have to work with big data in practice - small ones can be useful.
Science. The second component of data science. It's responsible for correct analyzing, finding useful regularities and their use. On this step it's useful disciplines like statistics, machine learning and optimization theory. They form the most important part of data science - data analyses. Machine learning helps to find regularities in historical data to forecast need information for new objects. But data science doesn't finish on founding the regularities in data. It's an applied research, where necessary to pay attention on hypothesis testing, experiment testing, result assessment and possibility for using in real cases. It's extremely useful for business-processes to understand if finding data science decision gives benefit for the project.
Let's consider methods and techniques for big data analyses. There are many different methods of analyzing data arrays, which are based on tools from statistics and computer science (for example, machine learning). The list does not pretend to be complete, however, it reflects the most popular approaches in various industries. Researchers continue to work on the creation of new techniques and improving existing ones.
In different researches are identified different methods and analysis techniques applicable to big data, but these lists aren't full. So, in this article it's represented wider variety of different methods and technics which have the possibility to be applied for Ukrainian businesses to raise their efficiency. Some of the represented techniques are not necessarily applicable exclusively to big data and can be successfully used for smaller arrays (for example, A / B testing, regression analysis). Of course, the more voluminous and diversified the array is analyzed, the more accurate and relevant data can be obtained at the output.
Data Mining Methods (business intelligence methods, deep learning analyses) - a set of methods for detecting previously unknown, non-trivial, practically useful knowledge necessary for making
decisions in data. They allow to determine the most susceptible categories of consumers for the product or service being promoted, identify the characteristics of the most successful employees, and predict the behavioral model of consumers. Such methods include association rule learning, classification (categorization), cluster analysis, regression analysis, detection and analysis of deviations, etc.
Association rule learning. A set of techniques for identifying relationships - associative rules between variables in large data sets. Used in data mining.
Classification - a set of techniques that allows to predict consumer behavior in a market segment (making purchasing decisions, outflow, consumption, etc.). Used in data mining.
Cluster analysis is statistical method for classifying objects into groups by identifying previously unknown common features.
Regression - a set of statistical methods for identifying patterns between changes in a dependent variable and one or more independent ones. Often used for forecasting and prediction.
Detection and analysis of deviations is the method which combines process mining with association rule mining to simplify the analysis of deviating cases. Association rule mining is used to group deviating cases into business rules according to similar attribute values. Consequently, only the resulting business rules need to be examined on their acceptability which makes the analysis less complicated [4]. Deviation analysis can reveal surprising facts hidden inside data.
Crowdsourcing - classification and enrichment of data by the forces of a wide, indefinite circle of people performing this work without entering an employment relationship. The technique of collecting data from many sources.
Data fusion and integration is a set of techniques that allow to integrate heterogeneous data from a variety of sources for the purpose of conducting in-depth analysis (for example, digital signal processing, natural language processing, etc.)
Machine learning (artificial intelligence) - a direction in computer science, which aims to create self-learning algorithms based on the analysis of empirical data. It includes learning with and without a teacher (supervised and unsupervised learning) using models based on statistical analysis to obtain complex forecasts.
Supervised learning. A set of techniques based on machine learning technologies that identify functional relationships in the analyzed data sets.
Unsupervised learning. A set of techniques based on machine learning technologies that reveal hidden functional relationships in the analyzed data sets.
Artificial neural networks, network analysis, optimization, including genetic algorithms (genetic algorithm - heuristic search algorithms used to solve optimization and modeling problems by randomly selecting, combining and varying the desired parameters using mechanisms like natural selection).
Pattern recognition - a set of techniques with elements of self-learning to predict the behavioral model of consumers.
Predictive analytics predictive modeling - a set of techniques that allow to create a mathematical model of a predetermined probable scenario of the development of events. For example, an analysis of the CRM-system database for possible conditions that subscribers will push to change the provider.
Simulation - a modeling the behavior of complex systems that is often used to predict, and study various scenarios in planning. It allows to build models describing the processes as they would be. Simulation can be considered as a kind of experimental tests.
Spatial analysis - a class of spatial data analysis methods, partly borrowed from statistics, that use topological, geometric and geographical information extracted from data, such as topology of the area, geographical coordinates, and geometry of objects. In these methods geoinformation systems (GIS) are often the source of big data.
Statistical analysis - methods of collecting, organizing, and interpreting data, including developing questionnaires and conducting experiments as well as further analysis of data. Analysis includes such methods as time series analysis, A / B testing (split testing - the marketing research method; when the control group of elements is compared with a set of test groups in which one or more indicators have been changed in order to find out which changes improve the target) others.
Time series analysis. A set of analysis methods of data sequences, borrowed from statistics and digital signal processing, that analyze repeating over time.
A / B testing. A technique in which a control sample is alternately compared with others. Thus,
it is possible to identify the optimal combination of indicators to achieve, for example, the best consumer response to a marketing offer. Big data allows for a huge number of iterations to obtain a statistically reliable result.
Visualization of analytical data - presentation of information in the form of drawings, diagrams, using interactive features and animations to obtain results, and to use as source data for further analysis. It is very important stage in the analysis of big data, allowing to present the most important results of the analysis in the most convenient way for perception and decision making.
Data fusion and data integration. A set of techniques that allows to analyze the comments of users of social networks and compare with the results of sales in real time.
Ensemble learning. In this method, many predicative models are used, due to which the quality of the forecasts is increased.
Genetic algorithms. In this technique, possible solutions are presented in the form of 'chromosomes', which can be combined and mutated. As in the process of natural evolution, the fittest individual survives.
Natural language processing (NLP) - a set of
techniques borrowed from computer science and linguistics for recognizing the natural language of a person.
Network analysis. A set of techniques for analyzing connections between nodes in networks. In social networks, it allows to analyze the relationship between individual users, companies, communities, etc.
Optimization - a set of numerical methods for the redesign of complex systems and processes to improve one or more indicators. It helps in making strategic decisions, for example, the composition of the product line brought to the market, investment analysis, etc.
Sentiment analysis. The methods for assessing consumer sentiment are based on technologies for recognizing a person's natural language. They allow to isolate from the general information flow messages related to the subject of interest (for example, a consumer product). Further evaluate the polarity of the judgment (positive or negative), the degree of emotionality, and others.
Signal processing. A set of techniques borrowed from radio engineering that pursues the
goal of recognizing a signal against a background of noise and its further analysis.
Digitalization is not only capable of creating new business processes, organizational structures, regulations, and new role models. It is designed to simplify the activities of companies. But, despite this, the introduction of digital technology into the company is a huge work and a rather long process. Digital transformation needs a comprehensive solution to business problems, together with correctly selected IT tools.
Implementation of new technologies and work with big data makes the business transform. The rapid development of digitalization in a few years will lead to the fact that all companies will retain and manage customers using digital technologies. According to McKinsey's Global Institute, Big data: The next frontier for innovation, competition, and productivity report, data has become as important a factor in production as labor and productive assets. By using big data, companies can gain tangible competitive advantages. Big Data technologies can be useful in solving the following tasks: market forecasting; marketing and sales optimization; product development; management decisions; increase in labor productivity; efficient logistics; monitoring the status of fixed assets and others.
References
1. Reports from Frost & Sullivan. URL: https://www. marketresearch.com/Frost-Sullivan-v383/
2. David Loshin. Big Data Analytics. From Strategic Planning to Enterprise Integration with Tools, Techniques, NoSQL, and Graph, 2O13, 142 p.
3. Steve Williams. Business Intelligence Strategy and Big Data Analytics. 1st Edition, 2O16, 24O p.
4. Swinnen J., Depaire B., Jans M.J., Vanhoof K. (2O12) A Process Deviation Analysis - A Case Study. In: Daniel F., Barkaoui K., Dustdar S. (eds) Business Process Management Workshops. BPM 2O11. Lecture Notes in Business Information Processing, vol 99. Springer, Berlin, Heidelberg.
Даш про автора Буркна На^ля Валерй'вна,
K.neftK, Донeцький нaцiонaльний y^Bepa/rreT ÍMe^ Bacиля Cryca, доцeнт кaфeдpи бiзнec-cтaтиcтики Ta eкономiчноï кiбepнeтики e-mail: nvburkina@gmail.com
Данные об авторе Буркина На^лья Валериевна,
K.neftK, Донeцкий нaционaльный yнивepcитeт имeни Bacыля Cryca, доцeнт кaфeдpы бизнec-cтaтиcтики и экономичecкой кибepнeтики e-mail: nvburkina@gmail.com
Information about the author Natalia Burkina,
Ph.D., Vasyl Stus Donetsk National University, Associate Professor, Department of Business Statistics and Economic Cybernetics e-mail: nvburkina@gmail.com
УДК 330.42:330.47:005.334-047.58-048.23 http://doi.org/10.5281/zenodo.3829628
КОЛОД1ЙЧУК А.В.
Економшо-математичне моделювання ризишв впровадження шформацшно-комушкацшних технологш
Предметом до^дження е економко-математичне моделювання ризик'1в '¡нформатизацИ господарських в'щносин на piBHi держави i perioHiB.
Метою до^дження е визначення параметрiв економко-математичного моделювання ризи-KiB впровадження сучасних iнформацiйно-комунiкацiйних технолопй.
Методи досл'!дження. У робот використан'1 д^алектичний метод наукового пзнання, метод анал'зу i синтезу, пор'1вняльний метод, метод узагальнення даних.
Результат роботи. В статт'1 видлено класи методов i моделей економко-математичного моделювання ризик'1в на макроеконом'чному рiвнi, зокрема методи елементарно'1 математики, методи математичного анал'зу i математично'1' статистики, методи математичного програмування та низку прикладених математичних теорй моделювання ризик'1в 1КТ. Придлено особливу увагу осо-бливостям застосування статистичних мето^в i моделей у вищеокресленй сферi.
