PROJECT TITLE :

Predictive Approach for User Long-Term Needs in Content-Based Image Suggestion

ABSTRACT:

In this paper, we formalize content-based image suggestion (CBIS) as a Bayesian prediction drawback. In CBIS, users offer the rating of pictures per both their long-term desires and also the contextual state of affairs, such as time and place, to that they belong. Thus, a CBIS model is defined to fit the distribution of the info so as to predict relevant pictures for a given user. Usually, CBIS becomes challenging when only a little amount of information is available like within the case of “new users” and “new pictures.” The Bayesian predictive approach is an efficient solution to such a drawback. Moreover, this approach offers economical means that to pick highly rated and diversified suggestions in conformance with theories in client psychology. Experiments on a real knowledge set show the deserves of our approach in terms of image suggestion accuracy and potency.