Impact of Domain Analysis on Reuse Methods

The Government is faced with an urgent need to increase productivity in the development and maintenance of mission critical computer systems. Reusing software across multiple application systems is one way to accomplish this goal. Recent research in the area of reusable software indicates that domain analysis is the first activity which should be performed during the development of reusable software. A domain analysis identifies commonalities between systems within a given problem domain. These commonalities (typically represented as objects, operations and relationships which characterize the domain -- a domain model) are then implemented as software components which can be reused by new systems within that domain. Few domain analyses have been done to date, largely due to the complexity of the problem and the expense of the process. Well- defined methods for performing domain analyses and for measuring their 'goodness' do not yet exist. The immaturity of formal methods is the probable cause for the almost total lack of domain analysis tools. The primary objective of this effort is to develop an approach to make domain analysis practical and effective for the development of reusable software and the reuse of that software in new application systems. The main focus is domain analysis within the context of DoD software development; more specifically, the interest is in embedded Ada software for Army applications.