Abstract
For private enterprises, it has become increasingly important and indispensable to utilize and purchase an advanced information system from a software provider. The system helps firms to keep an eye on items such as the volatile market conditions, competitor changes and changing consumer needs. Generally speaking, installation of the specific information system starts when both a private enterprise and a provider in coordination begin to discuss planning for the installation with a "Requirement for Proposal (RFP)" being issued. This process eventually proceeds to testing of the system before it is finally put into place. Later, software maintenance becomes an important issue. Recently, it has become evident that, for many private firms, the expenditure for the software maintenance seems increasingly more expensive than that of the original installation and development of the software. With this in mind, this paper attempts to focus on software maintenance and the problem thereof. This paper deals with two related issues. First, it discusses system maintenance and tries to come up with a model best suited for this purpose. The paper attempts to study system repatriation. When requiring system maintenance, a company usually wants to know the cost and the timing of the maintenance work. The information on these issues is estimated in this paper as "the additional dates of the system repatriation." A logistic curve is applied to evaluate the expenditure. The second issue this paper examines is development of a system to measure the approximate cost of repatriation. In developing this information, the paper highlights a decision-tree method. Taking advantage of these two different information sets, administrators in a given firm are given a rough idea as to when maintenance work starts and how much the costs of this process for the enterprise, thus allowing them to be well prepared for software maintenance.
