Automating Structured Decisions
Decisions that are simple and/or highly structured are good candidates for automation. A structured decision is one in which the inputs, decision criteria, method to process those inputs, and specific outputs are well-defined. For example, it is a relatively straightforward task to develop a computer program that will generate late payment notices for those customers who have not paid their monthly telephone bill on time. The program requires certain input data about billing information, decision rules about what constitutes a missed payment, and output instructions to generate a form letter and/or billing notice to be sent to the customer.
The automation of structured decisions is also used to improve communication and decision making between organizations. Not long ago, it was necessary for retailers to call credit card companies every time a customer wished to charge a purchase using a credit card. A representative working for the credit card company would examine the customer's record to determine whether or not to approve each purchase. Today, this process has been automated so that information and decisions related to purchases using a credit card are transmitted electronically.
Now consider the many benefits of automating this simple process. First, automating this process has reduced the cost of operations for credit card companies. Card issuers have streamlined and downsized their staff since fewer customer representatives are needed to handle purchase approvals/disapprovals. Second, the automated process has sped up the approval process so that retailers can deliver prompt and reliable service to their customers (improving customer satisfaction) and process more financial transactions within a given time period. Thirdly, the communication between retailers and card issuers has been improved. More purchases can be approved with greater accuracy using the IT-enabled system.
Automating Unstructured Decisions
Many situations involve more complex decisions called unstructured decisions. As you would expect, these decisions are much more difficult to automate. Management science is a field of research that seeks to impose structure on unstructured decision-making situations. The intent is to apply IT, mathematical models and high-power statistical analysis to support - though not replace - the decision-making process. For example, simulation spreadsheets are used to identify potential solutions by testing a variety of decision-making assumptions (e.g., deciding whether interest rates will go up vs. interest rates will go down based on stock prices, the value of the dollar, and inflation rates), and optimization models which are used to determine a "best" solution given a predefined set of resource constraints.
Cindy Tri Septiani / 2KA34 / 11111662
Sumber: The Pennsylvania State University,[online],http://www.personal.psu.edu/glh10/ist110/topic_old/topic10/topic10_03.html
Tidak ada komentar:
Posting Komentar