Athens: World Scientific and Engineering Academy and Society
摘要:
This paper propose a novel algorithm for mining closed frequent sequences, a scalable, condensed and lossless structure of complete frequent sequences that can be mined from a sequence database. This algorithm, FMCSP, has applied several optimization methods, such as equivalence class, to alleviate the needs of searching space and run time. In particular, since one of the main issues in this type of algorithms is the redundant generation of the closed sequences, hence, we propose an effective and memory saving methods, different from previous works, does not require the complete set of closed sequences to be residing in the memory.