Abstract:This thesis investigates the application of Search Based Software Engineering (SBSE) approach in the field of Software Project Management (SPM). With SBSE approaches, a pool of candidate solutions to an SPM problem is automatically generated and gradually evolved to be increasingly more desirable. The thesis is motivated by the observation from industrial practice that it is much more helpful to the project manager to provide insightful knowledge than exact solutions. We investigate whether SBSE approaches can aid the project managers in decision making by not only providing them with desirable solutions, but also illustrating insightful “what-if” scenarios during the phases of project initiation, planning and enactment. SBSE techniques can automatically “evolve” solutions to software requirement elicitation, project staffing and scheduling problems. However, the current state-of- the-art computer-aided software project management tools remain limited in several aspects. First, software requirement engineering is plagued by problems associated with unreliable estimates. The estimations made early are assumed to be accurate, but the projects are estimated and executed in an environment filled with uncertainties that may lead to delay or disruptions. Second, software project scheduling and staffing are two closely related problems that have been studied separately by most published research in the field of computer aided software project management, but software project managers are usually confronted with the complex trade-off and correlations of scheduling and staffing. Last, full attendance of required staff is usually assumed after the staff have been assigned to the project, but the execution of a project is subject to staff absences because of sickness and turnover, for example. This thesis makes the following main contributions: (1) Introducing an automated SBSE approach to Sensitivity Analysis for requirement elicitation, which helps to achieve more accurate estimations by directing extra estimation effort towards those error-sensitive requirements and budgets. (2) Demonstrating that Co-evolutionary approaches can simultaneously co-evolve solutions for both work package sequencing and project team sizing. The proposed approach to these two interrelated problems yields better results than random and single-population evolutionary algorithms. (3) Presenting co-evolutionary approaches that can guide the project manager to anticipate and ameliorate the impact of staff absence. (4) The investigations of seven sets of real world data on software requirement and software project plans reveal general insights as well as exceptions of our approach in practise. (5) The establishment of a tool that implements the above concepts. These contributions support the thesis that automated SBSE tools can be beneficial to solution generation, and most importantly, insightful knowledge for decision making in the practise of software project management.

Dynamic Adaptive Search Based Software Engineering

Search-based software engineering: Trends, techniques and applications

Dynamic adaptive harmony search algorithm for optimization

Search Based Software Engineering [Guest Editorial].

Search Based Software Project Management

Transformed Search Based Software Engineering: A New Paradigm of SBSE

"Sampling"' as a Baseline Optimizer for Search-based Software Engineering

Dynamic Adaptation of Software-defined Networks for IoT Systems: A Search-based Approach

SWE-Search: Enhancing Software Agents with Monte Carlo Tree Search and Iterative Refinement

Cooperative Co-Evolutionary Optimization Of Software Project Staff Assignments And Job Scheduling

Dynamic Optimization in Fast-Changing Environments Via Offline Evolutionary Search

Search Based Requirements Optimisation: Existing Work and Challenges

Some Recent Work on Multi-objective Approaches to Search-Based Software Engineering

Decision Support for Dynamic Adaptation of Business Systems Based on Feature Binding Analysis

Achievements, Open Problems and Challenges for Search Based Software Testing

Software Engineering for Self-Adaptive Systems: A Research Roadmap

Dynamic Quality-Diversity Search

Runtime Analysis of Search Heuristics on Software Engineering Problems.

Synergizing Domain Expertise with Self-Awareness in Software Systems: A Patternized Architecture Guideline

An Adaptive Optimization Technique for Dynamic Environments

Automated Adaptations to Dynamic Software Architectures by Using Autonomous Agents.