Identifying Effective Factors on Using Service Oriented Achitecture in E-Banking

n


INTRODUCTION
Nowadays organizations face with unstable competitive environment which is rapidly growing. These fluctuations are due to various factors, e.g. technological changes, shortening of product life-cycle and globalization od economy . In the past, each organization defined a distinctive system for each process that was called island system. These systems Lack of communication caused problems like heterogeneous and replicated data, lack of interaction among processes, etc. so, the idea of integrative systems was introduced (Atashak&Mahzade, 2011). Since a business system contains various relevant sub-systems, designing and implementation of systems is a time consuming process. But if implementation of business processes would be done under services in the coverage of SOA, these problems can be overcome (Basias et al, 2015). Regarding the importance of SOA and organization tendencies and boosting information volume, this architecture acceptance has significant importance in banking systems. Integrating is important in the success of banking, especially e-banking and banks are required to integrate IT infrastructures for preparing proper ebanking services. SOA is an advanced integrating paradigm that can support banks in solving problems (Riad et al, 2008). In spite of this, many relevant issues regarding employing SOA in e-banking still remains unknown. For various reasons, including lack of a comprehensive framework which encompass different viewpoints, many banks experience failure in reaching the benefits of integration in e-banking. It is worth mentioning that limited background about effective factors on utilizing SOA in e-banking and lack of a comprehensive framework that explain these factors has created a major research issue that requires more studies and investigations (Alghatani, 2015). Nowadays banks improve their services continuously and develop safe electronic banking services via new channels (Lee et al, 2010). For this reason, they seek for efficient techniques in system integration and services. During the last decade they had the tendency to use SOA. Therefore the goal of present paper is to detect and analyze effective factors on employing SOA in e-banking and prioritization among them for determining the most effective factor and providing a comprehensive framework of effective factors. In order to conduct this study, key issues like e-banking, organizational architecture and service based organization has been concentrated. For this purpose, The distinctive feature of present study in comparison to the past ones is in developing effective factors of SOA acceptance and adding new dimensions to SOA framework. It also helps banks in taking important decisions related to using this architecture, realizing benefits, risks, costs and relevant changes of using SOA in e-banking.
In the next section, research background will be referred and proposed framework will be introduced. Then type and research method and results of data analysis by using SPSS software and final framework and conclusion will be presented, respectively.

RESEARCH BACKGROUNG
Various definitions have been given regarding electronic banking, some of them include: concentration on new distribution channels, presenting customers with new modified services, and utilizing electronic commerce strategies (BahriNejad&Khalaj, 2011) providing customers with banking services by using safe interfaces without physical presence (Allame&Zare, 2011) customer usage of internet for organizing, experimenting or conducting changes in bank account or investing in banks and new and traditional banking services (MoshrefJavadi, 2012) to customers via electronic mutual communicational channels (Zeithmal, 2000).
Service oriented architecture is not a new concept and has developed since 90s. What is new is the capability of realizing its successful implementation. SOA solution for integrating communicational systems is to connect informational systems via web-service. Since late of 90s, in order to challenge interactivity of informational systems ,different approaches have been presented in which the most famous ones include peer to peer and central interface based integration (Riad et al , 2008). In peer to peer, data are needed for interaction between two systems which requires providing specific data path. Such an approach was too expensive and is hard to deal with and making changes was difficult. With the advent of SOA previous problems were solved and sectors were defined as services so as to be changeable easily. Major purposes of SOA include standardization and integrating platforms and technological infrastructures in an organization, improving inter-organization interactivity, reusing and flexibility of information services, more alignment of information technology with business (Lee et al, 2010).
Nowadays banks seek for efficient techniques for integrating systems and services. One promising solution for banks to communicate among IT solutions and solving integration plans related to e-banking is SOA (Uthria&Rabhi, 2009). By employing this new technique and technology, customers can transfer their request via a standard coordinator on the web to bank. In this way, customer's data will be collected in a comprehensive data base (Siadat&Hemati, 2014). limited background around effective factors of employing SOA in e-banking and lack of a comprehensive framework that explains them, has created a major research problem that requires more studies (Niknejad et  As it was indicated, although researches have been conducted on SOA, a few studies have been conducted on effective factors of using SOA in e-banking. This issue has caused limitations, including: banks have problems dealing with integrating issues, While e-banking operations cover an important section of banking operations, and it will be developed in future. For this purpose, in this study, SOA frameworks and models were investigated for detecting effective factors of employing SOA in e-banking and a decision framework was prepared for using SOA in e-banking according to an extensive review. In order to categorize the effective factors, past research categories were used. Analyzing models and frameworks related to employing SOA indicates that the factors and their categories are not identical among researchers (Lawler et al, 2009). In order to conduct this study, 118 factors presented in recent studies among present backgrounds that were effective in using SOA in e-banking were detected. In depth analysis, factors were reduced to 24 in order to be investigated in a real banking environment. As it was indicated, according to past studies, these factors can be categorized in 4 groups that include; business, technical, human, procedural. They are illustrated in the following table.

RESEARCH METHOD
This research is of applied form since it intends to develop a technology for applying in a specific field of banking.
Research method is based on data gathering and it considers subject contents, present conditions of available phenomenon, and is of descriptive type. In order to measure the research variables, concepts were investigated by selecting a sample representative of statistic population and distributing questionnaire. In the next step the results were generalized to the whole population. Therefore the present study, is survey research regarding descriptive research categories (Sarmad et al, 1383). In this research a heuristic research model has been used. In order to gather operational data and measure variables, questionnaire and Likert scale have been used respectively. Experimental data extracted from a case study related to SOA in e-banking, have been prioritized and have been analyzed to get experimental results. The goal of prioritization in this qualitative research is to boost validity and reliability of finings. The proposed framework regarding effective factors on using SOA in e-banking has been tested via a case study in e-banking.

Determination of Validity And Reliability of The Questionnaire
First, in order to investigate the extent to which available sample questions represent the entire community of possible questions, we addressed the validity of data using SPSS software. Also, we conducted the reliability test to examine the correlation between questions of the questionnaire. If the correlation between scores of the tests measuring a single character is high, the questionnaire has convergent validity. The existence of this correlation is essential to ensure that the test will measure what is to be measured. For convergent validity, average variance extracted (AVE) and composite reliability (CR) are calculated. For this purpose, the following relationship should be established between them:  To investigate the reliability of data, Cronbach's alpha coefficient is used which is one of the most commonly used methods of measuring reliability of questionnaires and determines how well questions correlate with each other. Obviously, the more Cronbach's alpha index is close to 1, the more questions are internally correlated and therefore homogeneous. Cronbach considers reliability coefficient of 45% low and coefficient of 75% moderate and acceptable. In this research, as shown in Table 5, Cronbach's alpha coefficients of all variables are more than 0.7 and therefore all variables are approved in terms of reliability. Average variance extracted (AVE) is always more than 0.5, so convergent validity is also confirmed. Moreover, composite reliability (CR) is larger than AVE in all factors, so all conditions of convergent validity are satisfied and the sample questions appropriately represent the population.

DATA ANALYSIS Descriptive Analysis of Research Variables
To conduct this research, we used 2 questionnaires to assess independent and dependent variables. The statistical population consists of 80 employees of a private bank in Iran among which 70 people have been randomly selected as the sample. Descriptive analysis of dependent variables of the research has been presented in Table 6 based on central parameters (mean, median, and mode) and dispersion parameters (standard deviation, variance, and range). Descriptive analysis of dependent variables of the research has been presented in Table 7 based on central parameters (mean, median, and mode) and dispersion parameters (standard deviation, variance, and range).   I n t e r n a t i o n a l J o u r n a l o f C o m p u t e r s a n d T e c h n o l o   According to this table, it is clear that 70 correct responses to all questions of the research have been collected. These tables were set in descending order from the most effective to the least effective factors. Also, among dependent variables, the highest average belongs to technical factors with the amount of 3.821 which is higher than the average of Likert scale. The range fluctuates between 1 and 4. The minimum amount 3.252, is related to procedural factors. Procedural factors own the highest standard deviation. Median and mode show that most respondents chose options 3 and 4 which mean average and high. In the table related to independent variables, 70 correct responses to all questions of the research have been collected. Furthermore, among independent variables, the greatest average belongs to system integration factor with the amount of 4.478 which is even higher than the large amount of Likert scale. The lowest amount is for characteristics of the organization 2.692, being lower than the average amount of Likert scale.

Final Conceptual Framework of The Research
In this section, given the analyses of the data from questionnaires conducted using SPSS, the variables identified were prioritized and final research model has been presented. Based on average of responses to dependent variables, technical factors possessed the highest percentage following by business factors, human factors, and procedural factors. Among independent variables, system integration factor accounts for the highest percentage and the average of responses to the factors cost, complexity, external pressures, and organization features are less than the average of Likert scale and should be removed from the ultimate model.
According to an interview with and a survey of experts in the case of removal of two factors from procedural factors, it would be better if the two factors organizational culture and change management, which were among procedural factors and the average of the responses related to them was higher than the average of Likert, were maintained in the model. In this regard, it was found out that because these two factors are dependent upon organizational management, it is better to consider them under the umbrella of management in the organization and therefore they can be a subset of the business factor. The final research model is shown in Figure 2.
Considering the findings, in order to use service-oriented architecture in e-banking, it is first necessary to examine the affecting as well as inhibiting factors and then to make the plans needed to remove them. According to the analyses performed in the present study, owners and managers of a business, first, need to identify and provide the technology required for using this architecture in their own organization. Then, they have to investigate the factors affecting their business and its process so that the inhibiting factors can be identified. Finally, before implementing this architecture, they should provide personnel with their requirements and direct organizational culture towards the use of SOA in order not to face employee resistance and not to suffer from fatigue and stress while working.

CONCLUSIONS
In this article, we investigated the need to integrate in e-banking and the role of service-oriented architecture in this integration. Banks are in search of the effects of SOA in e-banking and a comprehensive framework for the factors affecting its use will help us understand the benefits of SOA in e-banking. Without doubt, creating a comprehensive banking system and organizing the information are among the key elements to help banks with risk management, fighting with money laundering, quality control, creating an integrated marketing system, and other administrative issues (Siadat et al., 2015). The present literature shows that limited research has been done in this area. Lack of complete studies on the analysis of the factors influencing the use of SOA in e-banking led to lack of a comprehensive framework. This lack of a comprehensive framework does not allow organizations to gain full advantages of using SOA in e-banking and this is important. For this reason, we studied levels of e-banking and the use of SOA by focusing on identification of effective factors. Based on literature review findings, we recognized 118 effective factors and evaluated them, coming to the conclusion that all 118 factors are not unique. As a result, we proposed a conceptual framework based on 24 factors with the highest importance including: cost, strategy, target, return on investment, IT and business alignment, communications, awareness and support of management, time, process standardization, complexity, IT infrastructure, security, reusability, system integration, agility, efficiency, and flexibility of processes, resistance to change, fatigue, stress, and skills of employees. Each identified factor plays an important role during the use of SOA in e-banking. These factors are crucial for successful use of SOA and do not have the same value. The framework presented in practical area has been tested through a case study. The proposed conceptual framework provides new insights regarding identification and classification of the factors affecting the use of SOA in e-banking. The previous analysis suggests that there are several factors and classifications of factors related to the use of SOA. Hence, researchers do not share the same views. We clarified a lot of confusion in this area and proposed a framework that helps identify and clarify the factors influencing the use of SOA in ebanking. The case organization is a private bank and after a survey of its personnel regarding the use of SOA we came to the conclusion that dependent factors in terms of priority include: technical, business, human, and procedural factors, technical factors having the greatest effect. Also, independent factors in terms of priority include: system integration, process standardization, IT infrastructure, agility, efficiency, and flexibility of processes, employee skills, IT and business alignment, resistance to change, organizational culture, change management, awareness and support of senior management, return on investment, strategy, communications, security, target, time, fatigue, reusability, adaptability,  stress, external pressures, cost, complexity, organization features. The independent factors cost, complexity, external pressures, and organization features have the least effect so that they can be removed from the proposed model, resulting in 20 factors remained among which system integration factor is the most important one. We believe that the proposed framework will help banks when deciding on the use of SOA. The proposed framework can be employed as a strategic tool that gives competitive advantages to banks and may allow them to move faster and use integration solutions in a successful way that increases the realization of their interests. Considering the importance of e-banking and SOA, the proposed framework is expected to have important theoretical and practical implications.