Content-Based Searching in Group Communication Systems

The importance of real-time pattern recognition in streaming media is rapidly growing (Liu, Wang, & Chen, 1998). Extracting information from audio and video streams in an e-collaboration scenario is getting increasing relevance as the networking infrastructure develops. This development enables the use of rich media content. Shared archives of this kind of knowledge need tools for exploration, navigation and searching. As an example, to filter out redundant copies of an audio record, added by different members of an e-community, helps to keep the knowledge-base clean and compact. The work presented here focuses on algorithms used for content-sensitive searching in audio. The main problem of such algorithms is the optimal selection of the reference patterns applied in the recognition procedure. The proposed method is based on distance maximization. It is able to quickly choose the reference pattern to be used by the pattern recognition algorithm (Richly, Kozma, Kovács, & Hosszú, 2001). The article presents a novel approach of searching patterns in shared audio file storages such as peer-to-peer (P2P) based systems. The proposed method is based on the recognition of specific patterns in the audio contents extending the searching possibility from the description based model to the content based model. The presented method identified as EMESE (experimental media-stream recognizer) is an important component of a light-weight content-searching system which is suitable for the investigation of network-wide shared file storages. The efficiency of the proposed procedure is demonstrated in the article. Background