We hypothesize that an extension-with-conditioning of Dempster-Shafer theory is suitable for encoding uncertainty and ignorance in the Relational Model. We present a formal and well-motivated definition of conditioning, and show the spirit of the required change in the Relational Model and some results that then follow. It remains to be investigated whether these results are satisfactory.
|Number of pages||8|
|Publication status||Published - 6 Jun 2006|