Team SILEX : Supporting Interaction and Learning by Experience
|Display the presentation of the Team|
|Coordinator||Amélie Cordier||Assistant coordinator||Pierre-Antoine Champin|
|Presentation||Members||News||Publications||International relationships||Industrial relationships||Projects and Research conventions|
Our research questions are organized in three topics, closely inter-related, and application projects at the cross-road of those topics.
Our common approach consists in considering the user/machine couple as a single learning system, co-evolving according to the pursued activity. The observation of that co-evolving system, enabled by the modeling of activity traces, allows us to make original proposals in the field of knowledge construction, user assistance, system adaptation to the user, and usage analysis by the user. In the user/machine couple, the “machine” is now to be understood as the networked digital environment of the user, obviously involving the Web in our research.
Our research questions lead us to design methods, define models and setup tools that we evaluate. Hence we connect theoretical study with applications, in domains as diverse as knowledge management, human learning or user disabilities.
Topic 1 - Knowledge dynamics and traced experience
The goal of SILEX in this topic is to propose a dynamic engineering of knowledge, using the records left by individual and collective activity in the digital environment. Those records are organized into a representation of traced experience, a source of knowledge co-built and available to sustain the activity of the user/machine couple. The situations investigated by technology enhanced learning system (topic 2) and adaptive interactive systems (topic 3) are the main sources of inspiration for the study of knowledge dynamics and traced experience.
The notion of reflexive trace of activity is studied in relation with cognitive sciences, information and communication sciences, and more generally with the field of complex systems.
Facet A - Engineering of knowledge dynamics
This facet taps on available records of the traced experience to co-build knowledge that is contextual, revisable, sharable, adaptable. It is about dynamically building systems and associated models (of knowledge, documents, users...) for the purposes of assistance, human learning, diagnosis, explanation, active reading, etc. Engineering of knowledge dynamics is essentially interactive, since it must follow the dynamics of the knowledge-producing activity. The user joins the analysts and the designer as a primary actor in the process of building knowledge: she interprets activity with her knowledge, that can not be reduced to its machine-readable representation. On the other hand, the digital environment infers candidate pieces of knowledge from the records of traced experience. This constitutes a general process of interactive knowledge discovery.
The modeling of learning from traced experience can also contribute to fields of cognitive sciences (epistemology, cognition and development).
Facet B - Modeled trace and Trace-Base Management Systems
From a theoretical point of view, the notion of modeled trace (M-Trace) is proposed as a digitial object with specified properties, well defined models, and operators with a precise semantics. This allows us to consider systems dedicated to the storage and handling of those objects, in order to autonomously manage traced experience in a digital environment. We propose the notion of Trace-Base Management System (TBMS), with an architecture, services and protocols that are theoretically grounded, and implemented as an open platform.
Facet C - Episode based reasoning and traced experience based reasoning
We are studying the dynamics of the knowledge involved in the CBR (Case Based Reasoning) cycle, by integrating interactive learning to the retrieval and adaptation phases, into a general principle of revision. We take into account the dynamics by considering activity traces as the context of problem solving episodes. We propose to generalize this approach in a unifying framework, associating a TBMS to a generic assistant, reusing episodes from traced experiences. The primary application field is the exchange and sharing of experience in using the digital environment between very different users (in terms of roles, interaction modality, domains, etc.).
Topic 2 - Co-desing of situated TELS
Our work on TELS (Technology Enhanced Learning Systems) is about the co-design of situated TELS. By situated TELS we mean TELS that are adaptable to both the context and the specificities of the learner. The co-design is achieved by the different actors involved with the TELS: authors, tutors, learners (in an individual or collective context), but also the TELS itself.
We consider that the design of TEL adapted to their users is enhanced by the interaction of these actors and continues with assessments and uses (according to the methods developed in the C facet of topic 3 below), in a continuous improvement process. Our research interests take place among others in the context of e-learning, serious games and mobile learning. Our approach is generic, in connection with model driven engineering. We claim that this genericity is not opposed to specific needs: it makes it possible to develop tools that address these needs by their flexible and customizable nature.
Facet A - Co-design system-human of TELS
In an engineering approach of TELS, we are proposing an approach of co-design involving both systems and the people involved: designers, teachers and learners. This co-design can take the form of authors environments, tools for monitoring the activity, or work on reflexivity based on the activity traces of various actors (with models and tools developed in topic 1).
Facet B - Customizing EIAH
A part of our work is related to the customization cycle of TELS. For this purpose, we use the activity traces of learners (relying again on the results of the facet B of topic 1). These traces are modeled and interpreted in order to develop learner profiles. These profiles allow to implement mechanisms to customize different aspects of a TELS: scenarios, activities, features, feedback, interface, context of use.
Topic 3 - Interactive Adaptive Systems
This topic of the SILEX team is about the interaction between human and machine (HMI), both in its individual and collective dimension (Computer Assisted Collaborative Work). A key point is the definition of models and methods for designing interactive systems that are adaptable and adaptive. Acceptability, usability and utility are at the heart of the designed GUI.
In addition to conventional interfaces, we develop research on “innovative” interactions using mobile and wearable interfaces, mixed and augmented reality, ubiquitous and ambient computing.
Several application areas are affected by this work: TELS and the Serious Games (in synergy with topic 2), disability, audiovisual and documentary practices, but also industry, the city 2.0 or intelligent homes.
Facet A - Design of Interactive Systems
We study in this facet the methods, models and tools to design and implement interactive systems that are useful, usable and acceptable. We rely on user-centered design methods (including by using the results of topic 1 on activity traces) and a model-driven engineering.
Facet B - Adaptation and Accessibility
In this facet we work on the dynamic adaptation of HMI to the triplet user / device / environment, based on human models, activity traces (reusing the results of topic 1) and context modeling.
The management of disability situtations is an important field for this research.
Facet C - Evaluation and use
In this facet we study more particularly the evaluation and use of interactive systems. We are working on assessment methods using heterogeneous data: observations and activity traces (using models and tools defined in topic 1). This work is more specifically focused on the context of human learning, contributing to topic 2 - co-design of situated TELS.
Last update : 2015-01-08 15:04:16