Bringing artificial intelligence to the modeling world

 

 

The Modelia project

Modelia aims to improve software development practices by integrating artificial intelligence techniques in model-based software engineering.

k

Smart Modeling IDEs

Better productivity thanks to an assisted modeling experience. For instance, providing syntactic and semantic autocompletion, where the IDE is able to guess the next steps in a modeling project and suggests them back to the modeler.

Ontology-based Modeling Bots

Playing the role of a virtual modeling assistant. Based on previous models on the same domain available online or on a comparison between your current model and general ontologies, the bot could suggest missing properties on the model, recommend best practices or warn that some model aspects differ from the way they are typically modeled.

Human-like Code Generators

A cognified code generator would be able to learn the style and best practices of a particular organization and mimic them in the code it outputs. By learning from good code samples, the generator would be able to imitate the company’s best practices and style and maximize its chances to be accepted as a new “developer” for the company.

Self-Morphing Collaborative Modeling Tools

Automatic learning approaches can also be used within self-morphing modeling tool, able to adapt its interface, functions, behaviour (and even the expressiveness of the language offered) to the expertise of the tool user and to the context of the modeling problem addressed.

Our Team

Jordi Cabot

ICREA Research Profesor, UOC

Sébastien Gérard

Research Director, CEA

Robert Clarisó

Lecturer and Researcher, UOC

Gwendal Daniel

Postdoctoral Researcher, UOC

Loli Burgueño

Postdoctoral Researcher, UOC

Chokri Mraidha

Head of Laboratory, CEA

Shuai Li

Project manager and Researcher, CEA

Takwa Kochbati

PhD Student, CEA & UOC

Maxime Savary-Leblanc

PhD Student, University of Lille & CEA