Abstract: P-augmented matrix is proposed by adopting the structures and dynamic characteristics of packet sets to improve ordinary augmented matrix A^*. P-augmented matrix consists of internal P-augmented matrix A^(-overF) and outer P-augmented matrix A^F, denoted by(A^(-overF),A^F). Under some certain conditions, P-augmented matrix can be reduced into ordinary augmented matrix A^*. By using the structures and dynamic characteristics of P-augmented matrix, the research of dynamic data mining is carried out. Several relationships, theorems and criterion are obtained as follows: the relationships between dynamic data mining and P-augmented matrix; the internal P-augmented matrix decision theorem, the outer P-augmented matrix decision theorem and the P-augmented matrix decision theorem for data dynamic mining; the P-augmented matrix criterion for dynamic data mining. Finally, these results are verified by an application.

Key words: P-sets, dynamic data mining, mining criterion, decision theorem, P-augmented matrix