A Typical Operation Sequence Discovery Algorithm Based on Association Rule

With the deep application of computer aided process planning, a wealth of process data has been accumulated in the manufacturing enterprises. To capture the inheritable experience and knowledge about the process planning from the data, the association rule is applied to discovery the typical operation sequence (TOS). An association rule model mining the TOS was built. In the model, a process route was a transaction, and an operation was an item. Therefore, the operation sequence was the subset of items and transactions. Each TOS was regarded as a rule. Based on the model, an improved A priori algorithm was presented to mine the TOS. The algorithm includes six steps: 1) generating frequent operation set; 2) the join step: generating the frequent operation sequence candidate set; 3) the prune step: reducing operation sequence in the frequent operation sequence candidate set; 4) calculating the support of every operation sequence; 5) generating frequent operation sequence set; 6) terminating the algorithm and obtaining the TOS. Finally, an example mining the TOS was analyzed. The analysis result explains that the algorithm is effectively applied to discovering the TOS.