A Modern Non Candidate Approach for sequential pattern mining with Dynamic Minimum Support

Finding frequent patterns in data mining plays a significant role for finding the relational patterns. Data mining is also called knowledge discovery in several database including mobile databases and for heterogeneous environment. In this paper we proposed a modern non candidate approach for sequential pattern mining with dynamic minimum support. Our modern approach is divided into six parts. 1) Accept the dataset from the heterogeneous input set. 2) Generate Token Based on the character, we only generate posterior tokens. 3) Minimum support is entering by the user according to the need and place. 4) Find the frequent pattern which is according to the dynamic minimum support 5) Find associated member according to the token value 6) Find useful pattern after applying pruning. Our approach is not based on candidate key so it takes less time and memory in comparison to the previous algorithm. Second and main thing is the dynamic minimum support which gives us the flexibility to find the frequent pattern based on location and user requirement.