Strategies for positive and negative relevance feedback in imageretrieval

Bibtex entry :

@inproceedings { VG:MMM2000b,
    author = { Henning M{\"u}ller and Wolfgang M{\"u}ller and St{\'e}phane Marchand-Maillet and David McG. Squire },
    title = { Strategies for positive and negative relevance feedback in imageretrieval },
    booktitle = { Proceedings of the International Conference on Pattern Recognition(ICPR'2000) },
    pages = { 1043-1046 },
    year = { 2000 },
    editor = { A. Sanfeliu and J.J. Villanueva and M. Vanrell and R. Alquezar andJ.-O. Eklundh and Y. Aloimonos },
    volume = { 1 },
    series = { Computer Vision and Image Analysis },
    address = { Barcelona, Spain },
    month = { sep 3--8 },
    url = { },
    abstract = { Relevance feedback has been shown to be a very effective tool forenhancing retrieval results in text retrieval. In content-based imageretrieval it is more and more frequently used and very good resultshave been obtained. However, too much negative feedback may destroya query as good features get negative weightings. This paper comparesa variety of strategies for positive and negative feedback. The performanceevaluation of feedback algorithms is a hard problem. To solve this,we obtain judgments from several users and employ an automated feedbackscheme. We can then evaluate different techniques using the samejudgments. Using automated feedback, the ability of a system to adaptto the user s needs can be measured very effectively. Our study highlightsthe utility of negative feedback, especially over several feedbacksteps. },
    owner = { steph },
    timestamp = { 2008.05.04 },
    url1 = { },
    vgclass = { refpap },
    vgproject = { viper },

Keywords: machine learning, information geometry, data mining, Big Data, affective information retrieval (recherche d'information), information visualisation, content-based image and video retrieval (CBIR, CBR, CBVR, CBMR, CBMIR), information mining, classification, multimedia and multimodal information management, semantic web, knowledge base (RDF, OWL, XML, metadata, auto-annotation, description), multimodal information fusion