A Method for Extracting Classifying Tags from the Digital Multimedia Content using the Neural Network Technologies

Every year the volume of multimedia content produced and stored in the Internet is growing steadily. The number of users consuming this type of content is increasing rapidly. These facts inevitably lead to the need to analyze and process this type of information both for commercial and research purposes; however, large amounts of the accumulated data require automating the processes. One of the first tasks that arises is the primary classification. This article describes a study of the system architecture designed and developed to classify the multimedia data. In the developed system, the multimedia content is classified by extracting a set of classifying text tags from the original multimedia file. The results indicate that a system with the architecture similar to the one developed could be applied in solving the problem of analyzing the multimedia data for classification purposes.