Association rule mining finds interesting association or correlation relationships among a large set of data items has the potential that produce many patterns. In spite of using minimum support and confidence thresholds to help weed out or exclude the exploration of uninteresting rules, many rules that are not interesting to the user may still be produced. We develop intelligent data mining technique that generate and evaluate association rules by semantic approach like common sense. We provide new and interesting knowledge to users by post-processing of datamining. We define a Common-sense measure by similarity between association rules and common sense knowledge. This measure is based on the common sense knowledge network.