My thesis is realized in the Ozalid project designed to improve the quality of digital documents , especially OCRed documents (Optical Character Recognition ) by instrumenting the collaboration of human experts through an editing social network. In the context of this project, the aim of my thesis is firstly to instrument the Ozalid platform to collect interaction traces of social network users , and secondly , to implement mechanisms and tools needed to exploit these interaction traces inside assistance process.
Editing digital documents, and more specifically , editing scanned documents is not a trivial activity. It contains potentially complex tasks that require knowledge of several people in several areas. So , one of the research questions is to assist users in such environments . For this, the paradigm of TBR ( Traced Based Reasoning) is used to identify problems and help users. One of the current challenges is to retrieve user experiences and shares them via a specific interface.