ANR COOBRA, 2025 -- ongoing.
The ANR COOBRA project (Cooperation between sighted individuals and visually impaired Braille users for learning and working together) is a participatory research initiative aimed at improving inclusion for visually impaired (VI) individuals in education and work. It focuses on accessible scientific content and collaborative tools, leveraging the NatBraille editor, with key challenges including VI access to two-dimensional structures (e.g., equations, tables, graphs) and collaboration between VI and sighted users. My role as postdoctoral researcher involves exploring multimodal interactions for VI, supporting collaboration through co-designed interfaces and processes, and leading tasks such as literature reviews, development of research questions, and dissemination.
ANR MOBILES, 2023 -- 2025.
The ANR MOBILES project seeks to document, understand, and support the spatial practices and language learning experiences of international students in French higher education institutions, emphasizing learning opportunities from immersive stays enhanced by digital tools. Its objectives include analyzing students’ spatial practices to uncover contextual learning opportunities, designing a city mapping interface for quantitative and qualitative data exploration, and implementing user-participation-based recommendation systems to support learning goals, while developing methods for collecting, modeling, analyzing, and visualizing geolocated activity traces from sources like photographs and social networks. My role involves developing a prototype smartphone application to collect digital data such as geographic traces and annotations, provide cartographic visualizations and discussions, enable students to annotate traces for sharing integration-facilitating locations, and utilize enriched traces to generate interpretable recommendations for diversifying practices based on individual factors, ultimately aiding social integration and awareness of learning through spatial practices.
LearningDataLab, 2022 -- 2023.
The Learning Data Lab (LDL) is an innovative initiative aimed at revolutionizing educational experiences through learning analytics, by building a robust infrastructure to collect, process, and analyze data from sources such as MOOC platforms like FUN and edX, as well as institutional learning management systems. The project integrates the ELK stack (Elasticsearch, Logstash, Kibana) with Grafana for comprehensive data visualization, enabling real-time insights and interactive dashboards that draw from ELK, MySQL, and MongoDB databases to deliver actionable insights for educators and students. Its objectives include enhancing learning outcomes via predictive learning analytics (PLA) using machine learning algorithms to identify student performance patterns and support informed decision-making. My role involved contributing to the infrastructure development, integration of tools, and implementation of analytics features.
SAVAStudio, 2022 -- 2023.
SAVAStudio is an application designed to support the orchestration and monitoring of hybrid classes that integrate face-to-face and online teaching and learning activities, comprising a Teacher Module for scripting and monitoring and a Learner Module for intuitive interaction. The Teacher Module enables educators to define sequenced steps for hybrid learning scenarios, monitor learners’ progress, detect struggling participants, track interactions, and assess engagement and comprehension through features like overall scores and prioritized requests, while the Learner Module allows students to report progress, self-assess understanding via a Likert scale, request help, and use a ‘panic button’ for urgent assistance. The project aims to enhance hybrid learning quality by providing flexible tools for personalized support and effective class management. My role was as the developer of this comprehensive platform, designing and implementing both modules.
LADStudio, 2021 -- 2022.
LADStudio is a web-based tool designed to facilitate the creation and management of decision-oriented learning dashboards, enabling users to identify patterns, extract insights, and make data-driven decisions through interactive visualizations that support situational awareness and sensemaking. It assists in specifying and generating dashboards via a cyclical process involving a specification wizard with five sequential screens, from target use description to dashboard generation and export, while employing a component-based approach with reusable libraries of indicators, visualizations, panels, and views classified by awareness levels. The tool follows a three-tier architecture and uses modern technologies, including Grafana for rendering prototypes, allowing users to generate, personalize, and export dashboards without technical skills. My role involved developing LADStudio, including its architecture, component modeling, and generative design features to empower collaboration between designers and end-users.
PADLAD, 2020 -- 2022.
PADLAD is a collaborative toolkit designed to empower educational practitioners in the co-creation of learning data dashboards. Its primary objectives include enabling effective tracking and identification of indicators, detecting and troubleshooting problematic areas, aggregating and contextualizing relevant information for informed decision-making, and facilitating the prototyping of customized dashboards to improve educational practices. My role involved designing and implementing this toolkit as a solution for educators, focusing on collaborative features and user empowerment.
COREADA, 2014 -- 2019
CoREADA is a research project in Learning Analytics that analyzes learners’ content consumption traces from OpenClassrooms MOOC courses to provide authors and educators with insights for improving their courses. The primary objectives include detecting reading sessions and developing a dashboard that offers easy-to-use analytics tools to better understand learners’ reading behavior, ultimately enabling the creation of more effective and engaging learning experiences. My role involved contributing to the research by achieving significant results in reading session detection and designing and building the dashboard.
CHM, 2010 -- 2012
The CHM (Component-based Hypervideo Model) project introduces a new data model to formally describe hypervideo documents and their behaviors, revolutionizing interactions with multimedia through video-centric hypermedia for immersive audiovisual experiences. Its main objectives are to develop a comprehensive formal model for hypervideo documents that extends beyond conventional multimedia and hypermedia models, and to create a practical framework for designing web-oriented hypervideos, enabling innovative multimedia applications across various fields. My role was as the author and leader of this project, developing it as a sub-project of the Advene Project.
MediaStudio, 2007 -- 2010
MediaStudio is a research project focused on the development of advanced multimedia systems and tools for the production and delivery of rich, interactive, and adaptable content. The project aims to propose innovative models and approaches that can enhance the user experience and engagement with multimedia content. By leveraging cutting-edge technologies and design methodologies, MediaStudio 1&2 aims to provide solutions that can be applied to a broad range of domains such as education, entertainment, and advertising. Ultimately, the goal of the project is to create novel multimedia experiences that can facilitate knowledge transfer, enhance learning outcomes, and improve audience engagement.