Value-added treatment inference model for rule-based certainty knowledge

During various knowledge sources and expert comments in the knowledge base may lead to knowledge overlaps, conflicts or data size variations in the knowledge base, with wrong knowledge leads to wrong decisions. This study proposes using an O–A–RV structure to express rule-based knowledge, integrating conditional probability, vector matrix and artificial intelligence, and building a conditional probability knowledge similarity algorithm, so as to obtain a similarity matrix of knowledge and determines correlations among knowledge. Also proposed to use reliability factor theory to express knowledge conflicts, overlaps and data size variations. Based on knowledge correlations, a rule-based knowledge value-added treatment inference algorithm is set up to run value-added treatments, so that wrong decisions can be avoided.