Analysis of Big Data Processing Impact on Decision-Making in Commercial Systems
Students Name: Vovchanska Nataliia Petrivna
Qualification Level: magister
Speciality: Information Communication Networks
Institute: Institute of Telecommunications, Radioelectronics and Electronic Engineering
Mode of Study: part
Academic Year: 2023-2024 н.р.
Language of Defence: ukrainian
Abstract: A significant number of information technologies have recently been used in everyday life. Instead of performing complex calculations yourself, we can use ready-made software. Such actions as buying goods or ordering services are more convenient to perform with the help of online services. Without a doubt, information technology has significantly improved the quality of work in many industries. User service time has been shortened, and the list of services they offer has expanded. At first glance, it seems that the intelligent algorithms are just perfect and always work efficiently and flawlessly. Information systems still face many challenges that should be resolved as soon as possible. Processing large amounts of data is an important task when providing services to users. Information collected from various sources is quite often difficult to process, contains a lot of redundancy, and does not contain important information. Analysis of large volumes of data causes a load on computing devices, which negatively affects the performance of the system as a whole. When analyzing large arrays of information, it is advisable to pre-analyze it and determine the part that is most important for further processing. Thus, the load on the devices in the system is reduced, the time for calculating the result is reduced, and the quality of user service is improved. Intelligent analysis of data from end nodes is an important aspect of improving the efficiency of information systems for various purposes [1,2]. For the effective operation of commercial systems, it is necessary to constantly collect and analyze data on various performance parameters, evaluations given by users of products, statistics, etc. Usually, such data are used to determine the strategy of the system, the distribution of resources between different nodes, and the reduction of the risks of abnormal situations. The use of algorithms for intelligent processing of big data significantly simplifies the tasks of identifying patterns and important characteristics. Many problems in commercial information systems require quick and effective solutions. To ensure the proper quality of user service, the system status, and ratings of individual goods and services are constantly monitored, and options for possible improvement are determined. Thanks to a structured approach, it is possible to significantly increase the performance indicators of both global and local tasks. The exchange of information between different systems and their components facilitates faster finding of optimal solutions for frequently recurring problems. Instead of an independent determination of the way to solve the problem, already existing ones are used [3,4]. Decision-making systems are used in many industries to automate the management of complex structures. To assess the correctness of the decisions made, various parameters are used, such as the number of manufactured and sold products, profit, and reduction of work failures. Based on such parameters, the decision-making system can adjust and improve its own results, thereby self-learning. In large-scale commercial systems, the processes of determining the most optimal solutions are quite complex, since a significant amount of information is processed. The more various parameters are taken into account, the more accurate and effective the result. However, as mentioned above, it is necessary to preprocess and filter large arrays of input data to prevent system overload. When establishing a match between users and products, data on purchase history and ratings,both positive and negative, are useful. Due to this, the task of determining the strategy of operation of commercial systems is simplified [5,6]. Thus, the specifics of decision-making systems are a very relevant issue today. In commercial systems, the volume of information that needs to be collected and stored is constantly growing, moreover, at a rapid pace. Therefore, it is advisable to use machine and deep learning methods, which allow us to analyze large volumes of information faster, as well as to determine the necessary data properties with high accuracy. Modern information technologies contribute to the introduction of methods of intelligent data processing in various fields [7]. Study object - Decision making. Scope of research - Data processing technologies. Goal of research - Study of decision-making methods based on the big data processing. In the work, a study of decision-making methods based on the big data processing was carried out. The peculiarities of the commercial system’s functioning were determined. The user behavior prediction model for adaptive decision-making regarding service prices was proposed.