Multimedia Retrieval and Classification for Web Content

Bibtex entry :

@inproceedings { Kludas:FDIA07:MRCWC,
    author = { Jana Kludas },
    title = { Multimedia Retrieval and Classification for Web Content },
    booktitle = { Proceedings of BCS IRSG Symposium: Future Directions in InformationAccess (FDIA) },
    year = { 2007 },
    address = { Glasgow, Scotland },
    month = { August 28-29 },
    abstract = { *The population of the World Wide Web with media of all types suchas texts, images, videos and audio files in recent years raised theattractiveness of multimedia retrieval. With our work on the influenceof dependencies between modalities and features we investigate whythese approaches still do not perform convincingly better than plaintext search approaches when applied to large, noisy collections likeweb content, even though these approaches have more information attheir hands. This article suggests that, due to the size and noise,the modality’s dependencies necessary for efficient information fusionbecomes small and hard to exploit. Preliminary experiments with twomulti modal collections underpin this statement.* },
    url = { http://viper.unige.ch/documents/pdf/kludas2007-fdia.pdf },
}