RAIHAUTI Teiki

PhD studen
in Computer Science and Modeling

Team: MAGE

I am a PhD student at LEPSE and at the AGAP institute . I studied computer science, earning a bachelor's degree in computer science from the Faculty of Sciences in Montpellier, followed by a master's degree IMAGINA. I then worked as an engineer at INRAe in the UMR Eco&Sols, then at LEPSE.

Start of PhD: 1.10.2023

Subject :  Semantic and modular representation of crop models

Abstract
The use of crop models to predict crop performance and environmental impact is becoming widespread at all levels. Their use to reduce input use, adapt agriculture to climate change, diversify agro-systems, and preserve biodiversity leads to constant review of their formalisms and modeling of new processes.
We have recently developed the Crop2ML (Crop Modeling Meta Language) model representation and transformation system, which allows model components to be developed in accordance with FAIR principles. The current limitations are the difficulty of adapting components to the Crop2ML representation and the difficulty of composing them into operational modeling solutions.
The thesis project aims to overcome these limitations by using recent advances in large language models (GPT models, Mistral, etc.) to facilitate the interoperability of components that are heterogeneous in terms of programming language and software architecture.
By proposing a hybrid modeling system, the thesis will make an original contribution to addressing the challenges of ecological, climatic, and digital transitions. The project has an international scope through its inclusion in the AMEI (Agriculture Model Exchange Initiative) coordinated by my PhD supervisors.

Key words :  Crop models, LLM, model reuse, modeling, semantics, software engineering

Supervisor
UMR LEPSE, MAGE Team: Pierre MARTRE (PhD supervisor)
AGAP Institut, Phenomen Team: Christophe PRADAL (PhD co-supervisor) and Myriam ADAM (PhD co-supervisor)
UPR AIDA: Cyrille Ahmed MIDINGOYI (PhD co-supervisor)