Personalized Information Recommendation for Multimedia Archive Navigation

Enriching multimedia archives with related information is essential for improving the access to these archives, especially when users need additional information but do not know how or where to search for it. In this project, we propose to create a framework for personalized information recommendation for multimedia archives which utilizes the information present in the archives (audio, video and documents) to retrieve related content from a variety of repositories, including the archive itself, additional databases, and websites. These recommendations will help the user to better understand the recording and to navigate through the archive.

This four-month project will create user-oriented resources which will enable the design of a new system, based on an existing Idiap system adapted to real-world use cases. These resources will include: (1) an analysis of the state of the art, with a description of the possible options (modular interfaces, personalization techniques, databases, query engines) with indications of feasibility and cost; (2) a demonstrator for a scenario in which users consult a multimedia archive of tutorials, and receive suggestions for additional material (this will include a video of the demo); (3) an elicitation questionnaire for individual users or focus groups to capture user needs.

Application Area - Exploitation of rich multimedia archives, Information Interfaces and Presentation, Perceptive and Cognitive Systems
Sep 01, 2011
Dec 31, 2011