Encoding the Uncertain OR Gate for Weighted Model Counting in Possibility Theory

Uncertain logical gates facilitate the representation of knowledge in possibilistic networks. As the number of parent variables increases, the number of parameters required to specify Conditional Possibility Tables becomes prohibitively large. Uncertain logical gates enable the automatic computation of these parameters, thereby reducing the workload of experts. They provide a method for modeling logical behaviors to combine binary variables using operators such as OR, AND, NOT, etc., while also generalizing these operators to non-binary variables. In this research, we will first focus on the computation of the parameters for the uncertain OR connector before proposing encodings for its generalization to non-binary variables through the uncertain MAX connector. This connector can be encoded using clauses in conjunctive normal form, with evaluation performed through weighted model counting and queries. We will compare various approaches for encoding the uncertain OR within the framework of possibility theory.