From AtlanMod

Kelly Garcés

Who I am

I am Phd. student at AtlanMod team (EMN-INRIA) and at ASCOLA team (LINA UMR CNRS 6241-INRIA CRIBA). My PhD work is funded by the project FLFS.

My Phd. work deals with the interoperability problem. The more longstanding solution to bridge heterogeneity is to establish mappings between data. The number of algorithms to compute mappings has rapidly increasing. A challenge is to make such algorithms much more generic. I propose a language named AML (AtlanMod Matching Language) that allows specify mappings algorithms independent of data technical space (e.g., Ecore, MOF, OWL). AML instruments the fundamental behind the most recent matching algorithms: to assemble multiple components each one employing a particular matching technique. In runtime, an AML program is a chain of model transformations, each one implementing a matching technique. Inputs/outputs flow along the chain as models.

An AML application is Model Adaptation, i.e., to adapt models to their evolving metamodels. AML algorithms detect metamodel changes. Then, a program takes the changes to derive adaptation transformations.

I have contributed AML to Eclipse, its implementation is on top of ATL (Atlas Transformation Language) and AMW (Atlas Model Weaver). I encourage you to use AML and please let me know how the tool is doing!

Education

I have a Master in Software Construction (University of Los Andes). My master thesis was titled: "Managing variability in model-driven software product line". In proposing an approach for generating applications that share common and different features from models, metamodels and transformations. A product line illustrates the approach. This was developed using ATL and AMW.

Research interests

• Model Driven Engineering
• Model transformation
• Model weaving
• Model matching
• Software engineering practices