The application in data mining by integrating matter-element with fuzzy theory

Owing to the threshold setting in generating association rules, the dependencies and implicit meanings of products should be different if one considers the quantity of every item, for instance, the TID1[(A,9), (B,4)] and TID2[(A,2), (B,1)], to infer large item set ABC from item set AB from transaction database or market-basket-data. Hence, in order to improve the subjective of point view setting, after generated association rules, the matter-element transformation of Extincs is applied and the field of characteristic of matter-element description in Extenics is replaced with the quantity of every item of corresponding value of fuzzified, i.e. membership degree as fuzzy theory. According to this, the all related items including items A, B, and C can be operated by matter-element and created the fuzzy inference rules to derive a more objective association rule base. Additionally, the weighted average method is used for counting the inference value if there were same business transaction in accordance with the transaction list. Thus, the quantity of item A and B can be inferred if the large item set AB can be generated from association rules. From above, the predictions or suggestion works, like as products promotion or add-on services, can be decided by the manager. Finally, we show experimentally that the proposal model can be predicted the customer´s trends or extracted implicit information.