Matching and analysing conservation–restoration trajectories

The context of this work is an on-going project at the French National Library (BnF), which aims providing predictions documents physical state based on their conservation–restoration histories. A document can be either in a good and available to readers, or damaged unavailable them. As libraries may contain millions documents, manual monitoring analysis not realistic practice. We therefore propose analyse conservation histories order derive reliable state. To achieve goal, we introduce paper following contributions. First, representation history as trajectory, define its different types events. also trajectory matching process that computes similarity score between two trajectories considering terminological heterogeneity events, using ontological model represents domain experts knowledge. Second, provide identifies most representative sequences events deteriorated documents. Finally, prediction for process. present some experiments showing effectiveness well model.