An Improved Bi-Temporal Model Based On Time-Related Attributes And The Application Of Temporal Association Rule

With the development of data mining technology, temporal information is concerned increasingly. Bitemporal database which can reflect not only the historical information of the incidents treated with but also the temporal information of meta events in the system has attracted increasing attention. With the development of the bitemporal database, the technologies of model and index have been developing and are reaching maturity in bitemporal database.In most cases, all the attributes in a record are not completely relative to time, and more and more researches and data analysis works of the hot spot are precisely on those time-related attributes. Although the traditional bitemporal database model is related to the effective time and transaction time, its effective time and transaction time agreed by the Panel in the whole of the record are not bound to a specific attribute. So the users are unable to distinguish the time-related attributes from others, and the relation tables in database are changed on record as smallest unit. These will make against the operational efficiency. At the same time, some of certain common data analysis works may become more complex, not directed in the traditional models. For example, when we make some of the conclusions of decision-making, we will focus on the time-related attributes at a particular period of time of the changes, but not the whole record.For these reasons, an improved bitemporal relationship model is proposed in this paper. The new bitemporal relationship model splites up the previous bitemporal relationship table into two parts: the host table and the object table based on the time-related attributes. The two parts use an object code labeling attribute "hostNo" to relate each other. So we have a separation between time-related attributes and other attributes unrelated to time. This may make the data analysis works which are focused on time-related attributes more directed. At the same time, the operations of insertion and modification of data are accelerated and the frequency of the connection operations in query is reduced as far as possible. In addition, in this paper, we describe the find, update operations in detail for the new model, including: add a new host, modify attributes, delete a host and so on.In order to verify the superiority of this new model in data analysis area, this paper focuses on the application of temporal association rules in data mining. Firstly, we give out the concepts of common association rules and the classical Apriori arithmetic. Secondly, on these bases, we submit some appropriate methods to deal with the special circumstances. Finally, we describe how to operate the temporal association rules algorithm on this improved new model and we have an operation procedure in detail and give out the experiment data. The expriments show that the new model has superiority in the application areas.