SOFTWARE RELEASE PLANNING WITH TIME-DEPENDENT VALUE FUNCTIONS AND FLEXIBLE RELEASE DATES

摘要

Release planning is of key importance for incremental software product development. In conjunction with the value and the effort needed to implement features, decisions need to be made as to which features are offered in which releases. The value of features can vary over time depending on market conditions, competition, contractual constraints, and other concerns. Release dates and the specific features placed in releases need to be determined in a way that maximizes the overall value related to the investments made. The main contributions of the paper are (i) the formulation of value-driven release planning where we allow time time-dependent value functions to express the value of features potentially assigned to releases, (ii) time-dependent functions expressing the resource capacities needed for implementing features, (iii) a solution method using genetic algorithms to determine release plans allowing variation of the release dates within a pre-defined interval of feasibility, and (iv) providing a proof-of-concept for the proposed approach by running a case study.