Negative association rules describe the mutual exclusion of different items
which is also very important. In this thesis
a novel negative association rules mining method called GMNAR is proposed by introducing discernable matrixes concept and Genetic algorithm idea. And the experiment shows that it is effective and efficient.