Source author record

Andrew Ireland

Andrew Ireland appears in the imported research catalog. Authorship, coauthor and topic links are available while profile ownership is still unclaimed.

ResearcherUnclaimed source record

Catalog footprint

What is connected

2works
3topics
4close collaborators

Actions

Connect this record

Log in to claim

Research graph

See the researcher in context

Open full explorer

Inspect adjacent papers, topics, institutions and collaborators without losing the researcher page.

Building this map preview

BZPEER is loading the nearby papers, people, topics and institutions for this page.

Published work

2 published item(s)

preprint2016arXiv

Semi-Automated Design Space Exploration for Formal Modelling

Refinement based formal methods allow the modelling of systems through incremental steps via abstraction. Discovering the right levels of abstraction, formulating correct and meaningful invariants, and analysing faulty models are some of the challenges faced when using this technique. Here, we propose Design Space Exploration, an approach that aims to assist a designer by automatically providing high-level modelling guidance in real-time. More specifically, through the combination of common patterns of modelling with techniques from automated theory formation and automated reasoning, different design alternatives are explored and suitable models that deal with faults are proposed.

preprint2011arXiv

Discovery of Invariants through Automated Theory Formation

Refinement is a powerful mechanism for mastering the complexities that arise when formally modelling systems. Refinement also brings with it additional proof obligations -- requiring a developer to discover properties relating to their design decisions. With the goal of reducing this burden, we have investigated how a general purpose theory formation tool, HR, can be used to automate the discovery of such properties within the context of Event-B. Here we develop a heuristic approach to the automatic discovery of invariants and report upon a series of experiments that we undertook in order to evaluate our approach. The set of heuristics developed provides systematic guidance in tailoring HR for a given Event-B development. These heuristics are based upon proof-failure analysis, and have given rise to some promising results.