A Recommendation System for Intelligent Issue Management
Issue and change management is a process that ensures the monitoring and information processing ability of an organization to deal with uncertainty and risk. Efficient change management is known to be a critical success factor in the evolution of software systems. The main objective of the project is to provide a methodology for "Intelligent Change Management" and to design, implement and evaluate a customized version of this methodology for the industrial partner. This includes modeling and analysis of processes and artefacts involved in reacting to change requests. We will study the ripple effects of change request and their implications for the software system as well as for customers and actual users of the system. The effort is directed towards monitoring, improving and enhancing Brightsquit's evolutionary development and operations processes and tools. In the presence of continuous change requests and management of new issues, the targeted methodology is intended to integrate all the results received from comprehensive data analytics and to offer them as a holistic recommendation system.
More precisely, the technical objectives O1 to O6 are:
(O1) Descriptive and predictive analytics on the ripple effect of requirements change requests.
(O2) Two-sided software development: User involvement in development and exchange of data.
(O3) Development of process measurement and analytics for optimized release management.
(O4) Data analytics for detecting process and issue patterns and their usage for quality management.
(O5) Design and implementation of a prototype recommendation system integrating results from (O1) to (O4).
(O6) Case study evaluation of the effectiveness and efficiency of the proposed approach.
The effective management of emerging issues is a prerequisite for successful operations of software-based systems across various domains (software systems, finance, health care). The methodology and prototype implementation developed in this project is evaluated in two case studies. The objective is to demonstrate the effect of improved software quality characteristics and responsiveness to change on key health care quality indicators.