Risk-driven revision of requirements models

AXEL VAN LAMSWEERDE; JEFF KRAMER; DALAL ALRAJEH; SEBASTIAN UCHITEL

Austin

Conferencia; Proceedings of the 38th International Conference on Software Engineering,; 2016

Requirements incompleteness is often the result of unantic- ipated adverse conditions which prevent the software and its environment from behaving as expected. These condi- tions represent risks that can cause severe software failures. The identification and resolution of such risks is therefore a crucial step towards requirements completeness. Obstacle analysis is a goal-driven form of risk analysis that aims at detecting missing conditions that can obstruct goals from being satisfied in a given domain, and resolving them.This paper proposes an approach for automatically revis- ing goals that may be under-specified or (partially) wrong to resolve obstructions in a given domain. The approach deploys a learning-based revision methodology in which ob- structed goals in a goal model are iteratively revised from traces exemplifying obstruction and non-obstruction occur- rences. Our revision methodology computes domain-consis- tent, obstruction-free revisions that are automatically prop- agated to other goals in the model in order to preserve the correctness of goal models whilst guaranteeing minimal change to the original model. We present the formal foun- dations of our learning-based approach, and show that it preserves the properties of our formal framework. We vali- date it against the benchmarking case study of the London Ambulance Service.