A. Thongtae, S. Srisuk

In this paper, we introduce an algorithm for reusable uninteresting rules base on concept lattices, chi-square test statistics and odds ratio. Our proposed method tries to reduce the search space by scanning the attributes in the database rather than the transactions. We known that the number of transactions much larger than the number of attributes, hence the search space can be decreased evidently. In our proposed method the weak rules will be evaluated in order to reuse them as strong rules. Our system is evaluated with experiments on the same database in [12].