Incremental delivery using abstract data types and requirements clustering

It has been recognized that abstract data types (ADTs) are important in software development. Developers use ADTs to implement systems. Customers are usually not able to understand ADTs due to lack of technical knowledge. Thus, they may not easily envision (or rather, may not be concerned with) how the use of ADTs meet their requirements. This paper uses the concepts of ADTs to accomplish incremental delivery and develops a methodology to deliver data-dominant systems incrementally. An important outcome of this methodology is an algorithm to cluster requirements based on ADTs obtained from the Onion methodology. A case study demonstrating the methodology is presented.<>