Study On Temporal Knowledge Discovery In Database And Its Application

According to the demand and status quo in the field of knowledge discovery, thepaper researches temporal knowledge discovery in database(KDD) and discoveres theunderlying periodicity and association rules. It includes data preprocessing,periodicity and association rule. Firstly, data preprocessing in knowledge discovery isresearched relatively deeply, general preprocessing methods are introduced and an2efficient ~ merging for discretization is proposed; Secondly, to discover periodicity,two simple, efficient and anti-jamming trend discovery methods are established,including inertia method and smoothness method; Finally, Apriori algorithm isimproved in order to discover the efficient temporal association rules in database. Theapplication to a group of traffic temporal data shows that the research is useful andefficient.