A linear method for deviation detection in large databases

摘要: We describe the problem of finding deviations in large data bases. Normally, explicit information outside data, like integrity constraints or predefined patterns, is used for deviation detection. In contrast, we approach from inside using implicit redundancy data. We give a formal description and present linear algorithm detecting deviations. Our solution simulates mechanism familiar to human beings: after seeing series similar an element disturbing considered exception. also experimental results application this on real-life datasets showing its effectiveness.