Introducing a new algorithm for sequential sequences to compress them in the exploration process

Data mining is the process of discovering valid, new, useful and understandablepatterns. It can be said that data mining is a process in which data analysis toolsseek to discover patterns and relationships between existing data that may lead tothe extraction of new information from the database. Two examples of sequentialpattern are: "۸۰% of customers who "They buy TVs, they buy cameras on the sameday." "Every time Microsoft shares fall ۵%, IBM shares fall ۴% in three days." Thefirst model helps us to manage the store shelf well, and the second model helps thecompany to function properly in the economic crisis. In this research, we firstreviewed the pattern extraction algorithms and examined them based on thehistorical order in which the algorithms were presented. These algorithms aredivided into two categories based on Apriori and FP-Growth. Then we introducedthe two proposed algorithms GoKrimp and SeqKrimp using the principle of minimumminimum description length, ie useful patterns that compress the database themost. Then we introduced our proposed method and named it newKrimp andviewed the results on a standard database.

Keywords:

Data mining , pattern discovery , hybrid database , extraction algorithm , sub-sequence