Adaptive multimedia content personalization

Abstract

Modeling multimedia content by identifying semantically meaningful entities can be arduous because it is difficult to simulate human perception. However, by creating an algorithm to respond interactively to user preference, content-retrieval systems can become more efficient and easier to use. In this paper, we investigate adaptive relevance feedback algorithms for interactive multimedia content personalization. In particular two interesting scenarios are examined. The first uses a weighted cross correlation similarity measure for ranking multimedia data. The second exploits concepts of functional analysis to model the similarity measure as a non-linear function, the type of which is estimated by the users' preferences. The algorithms are computationally efficient and they can be recursively implemented.