The Research Of Sequential Patterns Dsicovery In Database

Knowledge discovery in database (KDD) is a rapidly emerging research field relevant to artificial intelligence and database system,and discovery of sequential patterns is an important field in the KDD research.This thesis reseaches the problem of mining sequential patterns in databases.The main context is as follows:(l)An efficient algorithm for mining sequential patterns is presentred via using Concept Lattice and adapting the classical algorithm to reduce the number of scanning transaction database.(2)A algorithm for maintaining the discovered sequential patterns after inserting and deleting some transactions is presented;Moreove,a algorithm is devised,which employs sampling techniques to estimate the difference between the sequential patterns in a database before and after it is updated, and the estimated difference can be used to determine whether we should update the discovered sequential patterns or not.(3)A algorithm for mining sequential patterns with item constraints is presented.(4)The discovered sequential patterns represent the most common subsequences with the database sequences and can be used to determine their similarity.the clustering for database sequences based on discovered sequential patterns is researched.