Nowadays the population of the French national Library (BnF) is compound of 77% of students, researchers or teachers. This public is mainly asking for expertise and influences the way all the public have ass to a document. Alternatively, the "curious" public can’t appropriate expert tools. Not in the expertise this public is asking for experience : a new searching system based on cultural appropriation has to be thought.
Underscore is an aleatory search engine based on the online National Library ressources. It offers an immersive, strange and diffused experience. It operates by handling different medias in order to sort cognitive information by visual. It permits a more actual access to forgotten culture and an uncomplexified access to expert contents.
The user sequences his wants by severals tools helping him to define his needs. Keywords or filters (time, rythm, ambiance, geography, century) ask a request to the search engine. Then underscore generate an immersive video with forgotten images, videos and sounds. After this video, a list is shown to the user with all the documents seen before. He has now the ability to create « univers » and sort data as he want to. Others features like balancing the ambiance of a media, starting a video from a specific document are available with the aim of discovering by iteration.
Project directors : Stéphane Villard & Patrick de Glo de Besses