«

»

Mutation-based Model Synthesis in Model Driven Engineering

by Sagar Sen, Benoit Baudry
Abstract:
With the increasing use of models for software development and the emergence of model-driven engineering, it has become important to build accurate and precise models that present certain characteristics. Model transformation testing is a domain that requires generating a large number of models that satisfy coverage properties (cover the code of the transformation or the structure of the metamodel). However, manually building a set of models to test a transformation is a tedious task and having an automatic technique to generate models from a metamodel would be very helpful. We investigate the synthesis of models based on plans. Each plan comprises of a sequence of model synthesis rules (or mutation operators) specified as graph grammar (GG) rules. These mutation operators are primitive GG rules , automatically obtained from any meta-model. Such plans can be evolved by various artificial intelligence techniques to generate useful models for different tasks including model transformation testing.
Reference:
Mutation-based Model Synthesis in Model Driven Engineering (Sagar Sen, Benoit Baudry), In Proceedings of the mutation’06 workshop at ISSRE’06, 2006.
Bibtex Entry:
@inproceedings{sen06,
	Abstract = {With the increasing use of models for software development and the
	emergence of model-driven engineering, it has become important to
	build accurate and precise models that present certain characteristics.
	Model transformation testing is a domain that requires generating
	a large number of models that satisfy coverage properties (cover
	the code of the transformation or the structure of the metamodel).
	However, manually building a set of models to test a transformation
	is a tedious task and having an automatic technique to generate models
	from a metamodel would be very helpful. We investigate the synthesis
	of models based on plans. Each plan comprises of a sequence of model
	synthesis rules (or mutation operators) specified as graph grammar
	(GG) rules. These mutation operators are primitive GG rules , automatically
	obtained from any meta-model. Such plans can be evolved by various
	artificial intelligence techniques to generate useful models for
	different tasks including model transformation testing.},
	Address = {Raleigh, NC, USA},
	keywords = {test, MDE},
	Author = {Sen, Sagar and Baudry, Benoit},
	Booktitle = {Proceedings of the mutation'06 workshop at ISSRE'06},
	Month = {November},
	Title = {Mutation-based Model Synthesis in Model Driven Engineering},
	url = {http://www.irisa.fr/triskell/perso_pro/bbaudry/publis/sen06.pdf},
	x-abbrv = {Mutation},
	X-Country = {US},
	X-International-Audience = {yes},
	X-Language = {EN},
	X-Proceedings = {yes},
	Year = {2006},
	Bdsk-Url-1 = {http://www.irisa.fr/triskell/publis/2006/sen06.pdf}}