CISP-Growth: A Contiguous Item Sequential Pattern Mining Algorithm with Application Level IO Patterns

The previous study of pattern discovery in storage systems focus on sequential pattern (SP) mining in lower level traces, but they donpsilat scale well to the application level. For patterns in application level are mostly composed of Contiguous Item Sequential Pattern (CISP) which are much simpler than SP, so itpsilas inefficient for the previous studies to mine CISP with clumsy SP mining algorithms. We propose a novel algorithm CISP-Growth which is more preferable for mining application level IO patterns. The CISP-Growth only scan the origin sequence in one-pass and make patterns grew among slices to avoid the inefficiency and information loss in C-Miner approach. The experiment result shows that the CISP-Growth outperforms C-Miner prominently in mining with real application IO traces and the simulation result also proves the effectiveness of CISP in application IO optimizations.