Economic science/ 1. Banks and banking system.

Olga Stepanenko, PhD

Dmytro Sharaesvskyi, research assistant

Vadym Hetman Kyiv National Economic University

Intelligent Decision Support Systems in Banking

Current economic environment is characterized by increasing globalization, rapid development of information technologies and concentration of highly technological products. In such conditions much more attention is devoted to the issues of innovative development, effective financial and information communications and building a knowledge economy.

Innovative type of economic development that focuses on the generation, dissemination and use of knowledge becomes the main factor for the competitiveness of the country and determines the potential development of the economic system as a whole and of its components. And the main component, which provides innovative development is a modern intelligent information technologies. This is emphasised in the report "Implementation of Innovation Policy in Ukraine" by Oleg Khymenko from the Department of Innovation Policy of the State Committee of Ukraine for Science, Innovation and Information at the International Conference on Capacity Building in the Commercialization and Protection of Intellectual Property, which was held in Moscow in 2010 [1].

The development of intelligent decision support systems (IDSS) has recently gained significant acceleration. This is because the weight of decision making in a market economy has been significantly increased. That happened because market economy is characterized by high level of uncertainty. The banking industry is not an exception. All banking operations require IT support, which could provide automatic and effective work on all stages of decision-making process taking into account the level of risk. IDSS provide analysts and decision making units (DMU) factual basis for the decision in interactive mode.

Intelligent data analysis techniques become more and more popular in banking decision support systems [2]. This trend has several reasons: growth in the amount of information needed for a decision making; rapidly changing environment; the need to eliminate uncertainties associated with the lack of information; growing importance of decisions made by decision making units; need in a consistent approach to decision making based on a limited set of criterias; need to implement methods of financial management, which prevent loss of funds.

Therefore, there is an obvious need for DSS which could cover the entire cycle of data analysis, preparation and decision implementation.

Literature analysis shows an increased interest in automating the known heuristic approaches and in the application of modern econometric models to decision making process[3, 4]. Decision making significantly relies on creativity, skills and intuition of decision making units. Computer support for such activities in Ukraine is now restricted to the use of various software, which only solves some problems and does not cover the entire cycle of decision-making. IDSSs which are currently used differ significantly in their focus, purposes and functional orientation.

Uncertainty is an integral part of decision-making process. These uncertainties are divided into three classes [3]: uncertainties associated with the lack of knowledge on issues where the decision is made, uncertainties associated with the inability to accurately take account of the reaction environment for committed action; inaccurate understanding of their goals by the person receiving the decision.

It is fundamentally impossible to take into account these uncertainties using optimization problem along with strictly established criteria. The only way to remove these uncertainties is still associated with the subjective expert evaluation.

Computer support for decision-making process is based on the formalization of methods for obtaining interim evaluations and decision-making process algorithm. Formalization of methods for obtaining interim evaluations and their evaluation is extremely difficult task. The process of formalization depends mainly on the degree of problem understanding and methods, which are used for formalization. Simple solution to this problem is to use algorithms that are based on a combination of computer technical analysis indicators [2], whose parameters were optimized on historical sample data according to the specified criteria. Implementations of these algorithms can be considered as a simplified model of DSS-generator, which is a package of data processing software (analysis, forecasting, modeling, etc.). It allows to create specialized information environment on the basis of heuristic rules which are designed to recognize patterns of the banking system. This imposes restrictions on problem solving process because the process of obtaining a decision should be understood by the DMU. An overview of the basic blocks of this system will help to determine the basic functions that must execute IDSS.

Process of the development of such system differs significantly from the conventional software development. The main issue is the informality of the tasks which leads to the need to modify the principles and methods of IDSS construction simultaneously with the process of increasing knowledge base about the functioning of the banking system. Therefore the concept called prototype is usually used.

At the first stage the prototype is built, which should meet two requirements: it must solve the task and its complexity should be low. This allows to determine the suitability of specific models and the need to develop a new prototype.

The second stage is the verbal description of the problem and the identification of tasks and subtasks that IDSS should potentially be able to solve. Also input variables that are available are identified.

On the third stage, the stage of conceptualization, the types of input data are defined and the subproblems of the entire problem are identified. Also the protocol of DMU’s actions is generated.

On the fourth stage, development stage, the prototype is actually build. Development of a prototype is actually the programming of its components. The purpose of prototype building is to confirm that the selected solutions and methods are suitable for solving the individual subtasks, and for decision making in general.

The next step is testing of the IDSS on various test examples. The necessity of any changes is also identified.

Therefore the IDSS should also include some knowledge model which could be used by management processes. Hence, the system in question belongs to a class of systems of semiotic type with the ability to adapt. This choice from the set of management procedures is done by the mean of an adapter. The continues modification of knowledge base is done by the interpreter. The above interaction management system and controlled object are realized by the certain set of information flows. And directly in the chief DMU there is circulation of information flows as a result of forecasting and analytical work, software and information modeling and information security.

According to the above mentioned theory, a generalized model of the IDSS can be described by the expression: where A - the active elements of the system, E - passive elements of the system, R - relationships between elements, P_s - a holistic process of the system as a set of parallel interacting processes P_a. At the same time the elements of IDSS are interconnected by the relationships that were defined above.

It should be mentioned that one of the main areas where banking IDSSs play significant role is the banking activities monitoring. It is performed based on the main indicators of bank activities. For this purposes additional information requirements were identified [5].

According to the recommendations given in [6], for the banking IDSS there is a need to create a database which should include structured information in accordance with regulations and standards of banking industry: controlled parameters of N directions of banking; gradation characteristics that describe documents in the area of banking activities; the results of N directions that may be lost; acceptable probability of failure and risk management impacts (resources, assets, rate); correcting impacts (resources, assets, rates).

The conducted survey of current IDSS systems with limited functions, IDSS systems under development and available input and output data for the banking industry has showed that it is possible to define typical composition of the IDSS’s knowledge base (Table 1).

Table 1

Components of the IDSS’s knowledge base

Name of the component (information object)	Content	Potential users (departments of the bank)
1	2	3
Information about the current condition of bank’s management units	Financial indicators Indicators of socio-political status Indicators of economic and technical conditions Indicators of natural and environmental conditions	Information and analytical services
Statistical information on the status of bank’s management units	Statistics, the generalized specifications, diagrams, etc.	Statistical services
Incoming and outgoing documents	Details and brief content of incoming documents Details and brief content of the output documents Information about the staff’s tasks and the implementation of these tasks	Management office
Information about planning of the current banking activities	Information about planned activities and their implementation in the divisions of the bank Information about planned activities of the top management	Office of the current activities
Information about bank’s personnel	Information about positions, salary, etc. Current information (holidays, orders, etc.)	Human resource department
Accounting information	Data on the bank’s budget Bank's loan portfolio Data on taxes Data on the demand for the resources by bank’s divisions Information about bank’s contracts and agreements Information about bank’s costs	Management accounting department
Information about users and architecture of bank’s LAN, EPS and global network	Information about users of bank’s LAN, EPS and global network Information about resources of bank’s LAN, EPS and global network Information about the use of resources Monitoring of the access to resources	Department of information technologies

In addition to the defined above components each IDSS has its specific internal knowledge base components, as well as specific vertical information flows dedicated only for the central bank.

IDSS which is developed according to the above mentioned principals is an effective tool for decision-making in banking, which is risk-weighted by the means of simulation scenarios using bank’s strategic planning, investment, development of credit strategy, optimizing the structure of bank capital, marketing strategies, exchange activities and other factors.

So we can conclude that banking IDSSs represent a new class of systems that currently are not developed both in theory and in practice. Therefore advisable to continue research in this area to ensure the efficiency of both individual banks and the banking system as a whole.

References:

1. Хименко О. Реализация инновационной политики в Украине/ Електронний режим доступу: http://www.unece.org/fileadmin/DAM/ceci/ppt_presentations/2010/ ip/Moscow/khymenko.pdf

2. Stepanenko O.P. Innowacyjne technologie zaradzania antykryzysowego/ Ramazanow S.K., Levasheva L.W., Stepanenko O.P., Tymaszowa L.A., Zakrewski J.J.; Pod red. prof. S.K. Ramazanowa. – Warszawa-Lugansk-Kijow: Reznikov V.S., 2011. – 368 s.

3. Э.А.Трахтенгерц Компьютерные методы реализации экономических и информационных управленческих решений. В 2-х т. Т.1. – М.:«Синтег», 2009. – 396с.

4. O.Stepanenko. Perspective Directions of the Banking System’s Stabilization/ О.Stepanenko// Perspektywiczne opracowania sa nauka i technikami – 2010. Materialy VI Miedzynarodowej naukowi-praktycznej konferencji. – Przemysl: Nauka I studia, 2010. – P. 20-23.

5. Степаненко О.П. ЛІ-моделювання та ІТ-підтримка процесів управління банківськими ризиками / О.П.Степаненко// Научный журнал «Культура народов Причерноморья». Крымский научный центр НАН и МОН Украина, ТНУ им. В.И.Вернадского, Межвузовский центр «Крым»./ – Симферополь, 2011. – № 205’2011 г.. – С. 218-222/

6. Daniel J. Power Decision Support Basics/ Daniel J. Power// Електронний ресурс. Режим доступу: http://businessexpertpress.com/books/decision-support-basics.