A case study to validate the PROXYMA approach to share and analyse contextualised interaction trace corpora in a TEL environment

Abstract

Analysing traces to personalise learning environments and study learning processes are common practices. Researchers usually define specific models for tracking the usage of learning environments. They also often develop specific analysis tools to study their specific research issues. The objective of this work is to define a framework called the 'Proxyma' approach intended for researchers to share contextualised interaction trace corpora as well as analysis tools despite their original specificity. To do so, we made the choice to avoid proposing a new pivot representation that has to be followed to make sharing possible. In fact, all existing approaches propose such solutions, which cannot answer all researchers' needs. Alternatively, we propose to share the data in their original representation and to define an intermediary layer called proxy which allows describing and querying the corpora. Our scientific process consists in attempting a first step of validation of the 'PROXYMA' approach by developing a system, as proof of concept, and by presenting an experiment using two external analysis tools.