| Summary: | 碩士 === 南台科技大學 === 資訊管理系 === 92 === Flourishing development of information science and technology, the pluralism that enterprises obtain the information channel more and more, obtain more and more materials too, including a lot of hidden information in these materials can be offered for formulation of the management tactics of enterprises'' reference, so must apply the technology prospected to the materials to prospect these hidden information to excavate out , so that enterprises can go to use these hiding information , but the way that is used most frequently in the materials are prospected is the related rule (Association Rules), because can find out the relation among the products and products from the related rule, having had this rule, enterprises can be regarded as the reference while stipulating the marketing tactics of the products according to this related relation found out of rule. Mostly prospect the database content of the regular content in the technology of prospecting of the materials that were studied in the past, but the databases of enterprises are that there are new materials to increase continuously every day in fact, if use the old related rule to prospect the way, though can prospect out and can offer enterprises as the information that is consulted , but must spend the operation time of very much one , do not accord with the economic benefits very much.
In this thesis, we have proposed the algorithms of performing of three improving types, can improve the efficiency of prospecting of the unusual fluctuation database , so the improvement type of a thesis performs algorithms and will not need to rescan the whole database each time when there are newly-increased materials, the majority only needs to scan the newly-increased trade materials, in order to improve the efficiency of prospecting.
|
|---|
