Organizations the world over face the challenging task of integrating multiple data sources in their quest to meaningfully and expeditiously exchange information among separately developed information systems architecture (Rosenthal & Seligman, 2001).
In their attempt to give users access to multiple data sources, these organizations must deal with a multiplicity of potential data integration issues in addition to settling the huge costs associated with initial implementation and continued maintenance of the data architecture (Tolk & Aaron, 2010). The present paper describes potential data integration issues for American Express.
There exist difficulties in integrating legacy applications into the main data architecture owing to the fact that many legacy applications rely on old or incompatible technology to store data (Tolk & Aaron, 2010).
Owing to its nature of operations and world-wide links, it would be plausible for American Express to convert legacy data to a “web format” using available applications (e.g., XML) and then integrate the data into company-wide Internet or Intranet infrastructure for effortless authoring, reclamation, updating and redistribution (Goethals, 2008).
For example, advertising and sales domain legacy data that may be important in determining the effectiveness of specific advertisements for the American Express company may be integrated into the information architect’s design by employing a tailor-made schema crosswalk or data warehouse to the converted data to enhance interoperability.
Potential data security issues may include how the company go about developing a privacy framework for data integration that is flexible and clear to end-users, how to develop schema matching solutions that safeguard the source data as well as schemas, and how to match entities and consolidate information about them across sources without necessarily revealing the sources or the real-world origin of the entities (Goethals, 2008).
In the advanced analytics domain, for example, the architect’s design systems may utilize access control or a set of privacy views under the privacy framework and have them coded in a declarative language extending SQL to prevent exposure of techniques used to identify thresholds for fraudulent activities that may be perpetuated by employees, customers or competitors.
In data implication, a major issue is how the architect would apply the suitable data federation and analytic strategies to multitudes of disparate data sources, with the view to extracting utilizable insights that could assist the company make better business and operational decisions (Rosenthal & Seligman, 2001).
In the accounts domain, for example, the architect may need to integrate customer feedback data with customer support information and firm’s performance, to extract value and competitive advantage from vast quantities of data collected.
The accounts, as well as advertising and sales domains, can be categorized as high-use domains within the American Express context. While the accounts domain deals with customer and human resource issues, qualifying it to be a high-use domain due to the nature, scope and importance of the data, the advertising and sales domain directly impacts on the company’s productivity and competitive advantage.
Consequently, the architect should integrate the data on the two domains by designing a system that allows users access to real-time data over a mediated schema or a virtual database (Goethals, 2008).
In integrating the high-use data domains, the systems architect must ensure that all the data sources are converted into interoperable formats in addition to developing and implementing standard service interfaces that readily interact with standard message layouts and transport protocols.
Lastly, IT governance issues within the American Express context include the creation of uncoordinated IT governance mechanisms, lack of senior management involvement in designing and implementing these mechanisms, lack of adequate incentive and reward systems, as well as lack of ownership and accountability for IT governance (Giachetti, 2004).
Consequently, the data integration system should be designed in a way that allows active senior management participation and regular reviews of each mechanism to ensure the focus remains on having the fewest number of effective mechanisms/domains possible.
References
Giachetti, R. E. (2004). A framework to review the information integration of the enterprise. International Journal of Production Research, 42(6), 1147-1166.
Goethals, F. G. (2008). Important issues for evaluating inter-organizational data integration configurations. Electronic Journal of Information Systems Evaluation, 11(3), 185-196.
Rosenthal, A., & Seligman, L. (2001). Scalability issues in data integration. Paper presented at AFCEA federal database colloquium, San Diego. Web.
Tolk, A., & Aaron, R. D. (2010). Addressing challenges of transferring explicit knowledge, information, and data in large heterogeneous organizations: A case example from a data-rich integration project at the U.S. army test and evaluation command. Engineering Management Journal, 22(2), 44-55.