Software Modelling and Evolution

Models are key to the automation of many software engineering processes, provide a level of abstraction that supports scalability and focus. Software evolution, including the reengineering, migration, refactoring, dynamic reconfiguration and adaption of systems, relies on reverse and forward engineering, analysis and redesign to be automated in order to be affordable and of high quality.

SOME is a research theme of the Department of Computer Science at the University of Leicester.

People

Academic staff

Ph.D. Students

Mohammed Alabdullatif

Abdullah Alshanqiti
Abdullah Alqahtani
James Hoey
Mudhaffar Hussein
Marwan Radwan

Overview

We use mainstream modelling languages and methods UML or BPMN and mathematical models such as automata, term and graph rewriting, process calculi, logics and semantic web languages and use their theory, techniques and tools, as well as transformations and mappings between them, to address software engineering problems.

Topics

Model transformations: Their use and in forward, reverse and re-engineering, refactoring and software migration.
Stochastic modelling and analysis: Based on stochastic graph transformations and associated model checking, simulation and approximation techniques we address performance of software systems and technical networks, and human behaviours in social networks.
Automated (model-based) testing and model inference: Verification is one of the most critical problems in software engineering and essential for software quality. In a model-based approach we focus on the relation between models and implementation (as in model-based testing) or extract models from implementation for further analysis (model inference).
Engineering of communication-centric software: in order to tackle the complexity of distributed and communication-centric applications it is natural to use models of distributed interactions. Recently, different models of such interactions have been studied in terms of high level specifications that capture the communication patterns of applications. The main objective of this research topic is to develop and integrate together approaches to express and analyse behavioural properties of distributed interactions.
Autonomous and adaptive systems: Service-oriented systems, cloud-based, mobile, and sensor-based applications (making up the Internet of Things) operate in highly dynamic environments where demands and resources change quickly. This blurs the boundary between design and runtime, requiring dynamic adaption and reconfiguration challenging safety, security and correctness. Performance requirements are high due to rapid collection and evaluation data.

Current Projects

2015 - present: DSTL ASUR EVIRE project (An Evidence-Based Reasoning Framework to Support the Transparent Control, Verification, and Validation of Autonomous Systems) - Walkinshaw PI

2015 - present: Department for Transport T-TRIG Grant: PREPAReD: Predicting, Preventing, and Analysing Rail Delays - Walkinshaw PI
2015 - 2017: KTP on Semantic Data Integration from Spreadsheets, with Synapse Information Ltd.
2015 - 2019: European COST Action on Reversible Computing

For more details or contacts, please drop Reiko an email.