Memory Island : Visualizing Hierarchical Knowledge as Insightful Islands. (Iles de mémoires : une nouvelle approche pour la visualisation intuitive des connaissances hiérarchiques)

This thesis is devoted to the study of an original cartographic visualization approach named Memory Island. We discuss how hierarchical knowledge can be meaningfully mapped and visualized as an insightful island. Our technique is inspired by the "loci" (plural of Latin "locus" for places or locations) method of the ancient "Art of Memory" technique. A well-designed map in mind can make sense of knowledge, which leads to the accomplishment of one's information seeking tasks, and helps to extend one's knowledge. To this end, Memory Island technique consists of associating each entity of knowledge to a designated area on a created virtual island. With the geographic visual metaphors we define, Memory Island can present phenomena found in knowledge, which is often difficult to understand. In this thesis, we discuss how we design our visualization technique to make it achieve the great features of visualization: automatically generate a truthful, functional, beautiful, insightful, and enlightening island with its technical details. In order to make Memory Island more convenient for its users, we present our "overview+detail" interface, to support them with visual exploration and knowledge analysis. We also demonstrate how to create knowledge maps using Memory Island technique, by giving some example on different datasets of Digital Humanities (Project OBVIL), e-books (Project LOCUPLETO) and other domains. Then, we propose our validation and evaluation protocols with two preliminary user experiments. The results from these studies indicate that the use of Memory Island provides advantages for non-experienced users tackling realistic browsing, helps them improve their performances in knowledge navigation and memorization tasks, and that most of them choose to use it for navigation and knowledge discovery. We end up by concluding our researches and listing some perspectives and future works that can be based on our Memory Island technique.