A Tool-Supported Methodology for Validation and Refinement of Early-Stage Domain Models

Model-driven engineering (MDE) promotes automated model transformations along the entire development process. Guaranteeing the quality of early models is essential for a successful application of MDE techniques and related tool-supported model refinements. Do these models properly reflect the requir...

Full description

Saved in:
Bibliographic Details
Published in:IEEE transactions on software engineering 2016-01, Vol.42 (1), p.2-25
Main Authors: Autili, Marco, Bertolino, Antonia, De Angelis, Guglielmo, Di Ruscio, Davide, Di Sandro, Alessio
Format: Article
Language:English
Subjects:
Automation
Biological system modeling
Context
Context modeling
Domain Modeling
Early Stage Model
Engines
Load modeling
Model Driven Engineering
Model Refinement
Model Validation
Natural language
Natural Language Questionnaires
Semantic Model Quality
Semantics
Software engineering
Software quality
Studies
Subject specialists
Unified modeling language
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Model-driven engineering (MDE) promotes automated model transformations along the entire development process. Guaranteeing the quality of early models is essential for a successful application of MDE techniques and related tool-supported model refinements. Do these models properly reflect the requirements elicited from the owners of the problem domain? Ultimately, this question needs to be asked to the domain experts. The problem is that a gap exists between the respective backgrounds of modeling experts and domain experts. MDE developers cannot show a model to the domain experts and simply ask them whether it is correct with respect to the requirements they had in mind. To facilitate their interaction and make such validation more systematic, we propose a methodology and a tool that derive a set of customizable questionnaires expressed in natural language from each model to be validated. Unexpected answers by domain experts help to identify those portions of the models requiring deeper attention. We illustrate the methodology and the current status of the developed tool MOTHIA, which can handle UML Use Case, Class, and Activity diagrams. We assess MOTHIA effectiveness in reducing the gap between domain and modeling experts, and in detecting modeling faults on the European Project CHOReOS.
ISSN:0098-5589
1939-3520
DOI:10.1109/TSE.2015.2449319