Learning operational requirements from goal models

Abstract

Goal-oriented methods have increasingly been recognised as an effective means for eliciting, elaborating, analysing and specifying software requirements. A key activity in these approaches is the elaboration of a correct and complete set of opertional requirements, in the form of pre-and trigger-conditions, that guarantee the system goals. Few existing approaches provide support for this crucial task and mainly rely on significant effort and expertise of the engineer.In this paper we propose a tool-based framework that combines model checking, inductive learning and scenarios for elaborating operational requirements from goal models. This is an iterative process that requires the engineer to identify positive and negative scenarios from counterexamples to the goals, generated using model checking, and to select operational requirements from suggestions computed by inductive learning.

On evidence preservation requirements for forensic-ready systems

“…More specifically, we first provide formal definitions ESEC/FSE'17, September [4][5][6][7][8]2017, Paderborn, Germany Dalal Alrajeh, Liliana Pasquale, and Bashar Nuseibeh of the domain model of a forensic-ready system including the environment in which incidents may occur and hypotheses about such incidents. We also formalise preservation specifications and requirements.…”

Learning and Representing Temporal Knowledge in Recurrent Networks

“…As a case study, we consider the problem of software model verification and adaptation, a successful application area of symbolic temporal logic. We have applied our model to the problem of verifying and evolving a specification of a water pump system [11]. The results indicate that neural-symbolic NARX networks can be used for both verification and learning, reducing the efforts involved in the modelling process and helping produce verifiable and sound system specifications.…”

“…The importance of adding learning mechanisms to temporal models has been highlighted in several applications, including model discovery and requirements acquisition in software engineering [7], [10], [11]. In what follows, we formally define a correspondence between recurrent networks and temporal logic.…”

“…In order to illustrate the different steps of the approach, we consider the pump system testbed used by [11]. The pump system monitors and controls the levels of water in a mine to avoid the risk of overflow.…”

“…In particular, the results corroborate the importance of adding background knowledge (when available) into neural networks learning. (3) The novel application of our methodology in the verification and revision of software models, providing an automated tool for the different processes as highlighted by [11], [41], [40], [42]; in addition, the integration with an existing model checking tool with several functionalities has led to results clearly useful in relevant application domains [38]. Limitations of the approach include, as discussed and analysed throughout the paper the difficulty in fully-automating the entire process, in particular the process of converting counterexamples into useful training sequences for learning.…”