Towards a version control model with uncertain data

• Authors: Mouhamadou Lamine Ba;Talel Abdessalem;Pierre Senellart
• Venue: Proceedings of the 4th workshop on Workshop for Ph.D. students in information & knowledge management
• Year: 2011

Content-based online collaborative platforms and office applications are widely used for collaborating and exchanging data, in particular in the form of XML-based electronic documents. Usually, a version control system is built-in in these applications to support collaboration and to properly manage document evolution. However, most version control models require error-prone and time-consuming manual management of conflicts and uncertainties, thus heavily affecting collaborative work. Since in collaborative contexts conflicts are of a more semantic nature, i.e., due to uncertainty on what is really true, we focus on a version control model that deals with uncertain data. We introduce an XML version control model to assess uncertainty in data and to automatically resolve conflicts. Office documents and those of content-based online collaborative platforms are described in XML. We give features and requirements of our targeted model by identifying important characteristics of versioning systems, in particular XML version control systems. As first results, we propose a version-space formalism for collaborative data and we put forward an uncertainty management approach by translating the problem in the framework of XML data integration.