• DocumentCode
    477011
  • Title

    Recombinant Cognition Synthesis - An introduction

  • Author

    Solano, Marco Antonio ; Tanik, Murat M.

  • Author_Institution
    Intell. & Inf. Syst., Raytheon, Dallas, TX
  • fYear
    2008
  • fDate
    June 30 2008-July 3 2008
  • Firstpage
    1
  • Lastpage
    8
  • Abstract
    Recombinant cognition synthesis (RCS) provides an architectural framework and methodology for combining multi-source metadata with predictive and analytic algorithms to synthesize actionable knowledge. RCS bridges an implementation gap in the current Joint Directors of Laboratories (JDL) data fusion model by specifically addressing the integration of fusion levels 2 to 4. The situation awareness (SA) envelope is predicated on a three-dimensional vector space, which comprises spatial, temporal, and semantic attributes of multi-source object metadata. The threat assessment (TA) envelope is defined by the vector space, which comprises means, motive, and opportunity weighted coefficients. Machine learning algorithms, statistical analysis, and predictive data mining techniques are applied to recombined metadata fused across its trans-dimensional attributes. An extension to the ISO 19115 model captures the necessary metadata for this fusion framework, and additionally provides the qualitative and probabilistic attributes to enable process refinement (PR).
  • Keywords
    data mining; learning (artificial intelligence); statistical analysis; data fusion; data mining; machine learning; multisource metadata; recombinant cognition synthesis; situation awareness; statistical analysis; threat assessment; vector space; Data Mining; Decision Support; Fusion Architecture; Fusion Levels 2, 3, 4.; GIS; JDL Fusion Model; Knowledge Discovery; Machine Learning; Multi-Modal/Source Data; Predictive Analytics; Situation Awareness; Threat Impact Assessment;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Information Fusion, 2008 11th International Conference on
  • Conference_Location
    Cologne
  • Print_ISBN
    978-3-8007-3092-6
  • Electronic_ISBN
    978-3-00-024883-2
  • Type

    conf

  • Filename
    4632396