Project-Team Flowers
Members
Overall Objectives
Research Program
Application Domains
Highlights of the Year
New Software and Platforms
New Results
Bilateral Contracts and Grants with Industry
Partnerships and Cooperations
Dissemination
Bibliography
XML PDF e-pub
PDF e-Pub

previous Previous | up Home | Next next

Section: New Results

Applications for Educational Technologies

Multi-Armed Bandits for Adaptive Personalization in Intelligent Tutoring Systems

Participants : Manuel Lopes [correspondant] , Pierre-Yves Oudeyer, Didier Roy, Alexandra Delmas, Benjamin Clement.
The Kidlearn project

Kidlearn is a research project studying how machine learning can be applied to intelligent tutoring systems. It aims at developing methodologies and software which adaptively personalize sequences of learning activities to the particularities of each individual student. Our systems aim at proposing to the student the right activity at the right time, maximizing concurrently his learning progress and its motivation. In addition to contributing to the efficiency of learning and motivation, the approach is also made to reduce the time needed to design ITS systems.

We present an approach to Intelligent Tutoring Systems which adaptively personalizes sequences of learning activities to maximize skills acquired by students, taking into account the limited time and motivational resources. At a given point in time, the system proposes to the students the activity which makes them progress faster. We introduce two algorithms that rely on the empirical estimation of the learning progress, RiARiT that uses information about the difficulty of each exercise and ZPDES that uses much less knowledge about the problem.

The system is based on the combination of three approaches. First, it leverages recent models of intrinsically motivated learning by transposing them to active teaching, relying on empirical estimation of learning progress provided by specific activities to particular students. Second, it uses state-of-the-art Multi-Arm Bandit (MAB) techniques to efficiently manage the exploration/exploitation challenge of this optimization process. Third, it leverages expert knowledge to constrain and bootstrap initial exploration of the MAB, while requiring only coarse guidance information of the expert and allowing the system to deal with didactic gaps in its knowledge. The system is evaluated in a scenario where 7-8 year old schoolchildren learn how to decompose numbers while manipulating money. Systematic experiments are presented with simulated students, followed by results of a user study across a population of 400 school children. [14]

A Comparison of Automatic Teaching Strategies for Heterogeneous Student Populations

Online planning of good teaching sequences has the potential to provide a truly personalized teaching experience with a huge impact on the motivation and learning of students. In this work we compare two main approaches to achieve such a goal, POMDPs that can find an optimal long-term path, and Multi-armed bandits that optimize policies locally and greedily but that are computationally more efficient while requiring a simpler learner model. Even with the availability of data from several tutoring systems, it is never possible to have a highly accurate student model or one that is tuned for each particular student. We study what is the impact of the quality of the student model on the final results obtained with the two algorithms. Our hypothesis is that the higher flexibility of multi-armed bandits in terms of the complexity and precision of the student model will compensate for the lack of longer term planning featured in POMDPs. We present several simulated results showing the limits and robustness of each approach and a comparison of heterogeneous populations of students.

Figure 17. ZPDES exploration of an activity graph, with δZDP the success rate over all active activities, λZPD the threshold to expand the ZPD, δAx the success rate for the activity Ax, and λa the threshold to reach to deactivate an activity.
IMG/ZPDESgraph.png

This work has been publish and presented at Educational Data Mining 2016 conference in Raleigh, USA [76].

Github link of the experiments paper code : https://github.com/flowersteam/kidlearn/tree/edm2016

The KidBreath project

To create learning contents linked to asthma to personalize it like mathematics activities in Kidlearn project [14] we used recommandation criterias in Therapeutic Education Program for asthma kids made by Health High Autority. Following an approach of participatory design [114], contents were validated by medical experts like health educators, pulmonoligists and pediatrics. Then, we conducted a workshop with forty kids aged 8 in order to iterate over the application interfaces and evaluate enjoy about it with observations. Finally, we realized a focus group with 5 asthma kids to validate the global comprehension of a part of the content. It revealed that children wanted more contents about the crisis treatment and how the asthma works in the human system (verbatims).

In a preliminary study, we experimented two conditions in 20 control children (with 3 asthma kids), one giving the possibility of choosing activities like the child wants, and one no giving this choice (activities displayed in random). No significative difference between the two groups, but results showed KidBreath was easy to use with scores > 75 using System Usability Scale [115]. Based on Cordova and Lepper works to evalueate motivation and knowledge with similar system and population [121], children had their disease knowledge increased with just one week use and were motivated using it. Finally, asthma kids showed they were more engaged than healthy kids and used KidBreath more seriously (stayed in breaks). These results was presented in the 5th edition of Serious Games in Medicine Conference in Nice.

We presented Thesis project in some events this year, with one publication surbmitted and validated:

Poppy Education: Designing and Evaluating Educational Robotics Kits

Participants : Pierre-Yves Oudeyer [correspondant] , Didier Roy, Théo Segonds, Stéphanie Noirpoudre, Marie Demangeat, Thibault Desprez, Matthieu Lapeyre, Pierre Rouanet, Nicolas Rabault.

The Poppy Education project aims to create, evaluate and disseminate all-inclusive pedagogical kits, open-source and low cost, for teaching computer science and robotics.

It is designed to help young people to take ownership with concepts and technologies of the digital world, and provide the tools they need to allow them to become actors of this world, with a considerable socio-economic potential. It is carried out in collaboration with teachers and several official french structures (French National Education, High schools, engineer schools, ... ). For secondary education and higher education, scientific literacy centers, Fablabs.

Poppy Education is based on the robotic platform poppy (open-source platform for the creation, use and sharing of interactive 3D printed robots), including:
Pedagogical experimentations : Design and experiment robots and the pedagogical activities in classroom

This project is user centred design. The pedagogical tools of the project (real and virtual robots, pedagogical activities, etc.) are being created directly with the users and evaluated in real life by experiments. For our experimentations in the classroom we are mainly using the robot Poppy Ergo Jr (real and virutal) and Snap! Our purpose is to improve this pedagogical tools and to create pedagogical activities and resources for teachers.

Figure 18. Experiment robots and pedagogical activities in classroom
IMG/robotics_in_classroom.jpg
Partnership on education projects
Created pedagogical documents and resources
Scientific mediation

To promote educational uses of the platform, we participated in events (conference, seminar etc.).

We participated as well at some workshops to introduce students to robotics and programming.
Symposium robotics

We organized a symposium robotics (http://dm1r.fr/colloque-robotique-education/) that present research results and feedback on the use of Poppy and Thymio robots in education (other robots have been discussed, such as BeeBot and Metabot), from kindergarten to higher education, The Centers for Scientific and Technical Culture.

It was a 2 day event : 200 participants, 40 speakers (conferences and workshops).

Poppy Education team and the working group teachers helped with the organisation of the event and during the event (talk and workshops).

All conference videos are available on the web :

https://www.youtube.com/watch?v=prFmC-BpdY8&index=1&list=PL9T8000j7sJBC_H3L_hS-i4Ltlh1Fz2FY

IniRobot: Educational Robotics in Primary Schools

Participants : Didier Roy [correspondant] , Pierre-Yves Oudeyer.

IniRobot (a project done in collaboration with EPFL/Mobsya) aims to create, evaluate and disseminate a pedagogical kit which uses Thymio robot, open-source and low cost, for teaching computer science and robotics.

IniRobot Project consists to produce and diffuse a pedagogical kit for teachers and animators, to help to train them directly or by the way of external structures. The aim of the kit is to initiate children to computer science and robotics. The kit provides a micro-world for learning, and takes an enquiry-based educational approach, where kids are led to construct their understanding through practicing an active investigation methodology within teams. It is based on the use of the Thymio II robotic platform. More details about this projects were published in RIE 2015 [50] , which presents the detailed pedagogical objectives and a first measure of results showing that children acquired several robotics-related concepts. See also http://www.inirobot.fr.

Deployment: After 24 months of activity, IniRobot is used by about 1400 adults and 16 000 children in 54 cities of France. Example of action in university: MEEF teacher training for the hope of Aquitaine. Example of action in school: training of all Gironde Pedagogical ICT Advisors, covering nearly 1000 schools. Example of action in the extracurricular time: training 82 facilitators TAP cities of Talence, Pessac, Lille, ..., CDC Gates of inter-seas. Example of national action: Training of the digital mediators of the 8 Inria centers.

Partnership

The project is carried out in main collaboration with the LSRO Laboratory from EPFL (Lausanne) and others collaborations with French National Education/Rectorat d'Aquitaine, with Canopé Educational Network, with ESPE (teacher's school) Aquitaine, ESPE Martinique, ESPE Poitiers, National Directorate of Digital Education
Created pedagogical documents and resources
Scientific mediation

Inirobot is very popular and often presented in events (conferences, workshops, ...) by us and by others.
Symposium robotics

With Poppy Education, Inirobot is a main line in our colloquium "Robotics and Education" (http://dm1r.fr/colloque-robotique-education/)


previous Previous | up Home | Next next