Towards statistical comparison and analysis of models

Ö. Babur, L. Cleophas, T. Verhoeff, M. van den Brand

Research output: Chapter in Book/Report/Conference proceedingConference paperAcademicpeer-review

11 Citations (Scopus)

Abstract

Model comparison is an important challenge in model-driven engineering, with many application areas such as model versioning and domain model recovery. There are numerous techniques that address this challenge in the literature, ranging from graph-based to linguistic ones. Most of these involve pairwise comparison, which might work, e.g. for model versioning with a small number of models to consider. However, they mostly ignore the case where there is a large number of models to compare, such as in common domain model/metamodel recovery from multiple models. In this paper we present a generic approach for model comparison and analysis as an exploratory first step for model recovery. We propose representing models in vector space model, and applying clustering techniques to compare and analyse a large set of models. We demonstrate our approach on a synthetic dataset of models generated via genetic algorithms.
Original languageEnglish
Title of host publicationIn Proceedings of the 4th International Conference on Model-Driven Engineering and Software Development - MODELSWARD
Pages361-367
Number of pages7
DOIs
Publication statusPublished - 2016
Externally publishedYes
Event4th International Conference on Model-Driven Engineering and Software Development, MODELSWARD 2016 - Rome, Italy
Duration: 19 Feb 201621 Feb 2016

Conference

Conference4th International Conference on Model-Driven Engineering and Software Development, MODELSWARD 2016
Country/TerritoryItaly
CityRome
Period19/02/1621/02/16

Keywords

  • Clustering
  • Model Comparison
  • Model-Driven Engineering
  • R
  • Statistical Analysis
  • Vector Space Model

Fingerprint

Dive into the research topics of 'Towards statistical comparison and analysis of models'. Together they form a unique fingerprint.

Cite this