Adaptive Discovery of Indexing Rules for Video

This paper presents results, at an early stage of research work, of the use of fuzzy decision trees in a multimedia framework. We present the discovery of rules in three different indexing scenarios. These rules represent knowledge that can be interpreted as guidelines for the development of better indexing tools. We use a fuzzy decision tree algorithm to extract these rules (just) from color proportions of key-frames extracted from one video-news broadcast. Experimental results and comparisons with other data mining tools are presented.