Pattern Mining

The term pattern is used in many different fields and has various meanings. In data mining, it usually refers to finding reoccurring structures in the data such as itemsets, subgraphs, or sequences. A typical example, which is often used to motivate pattern mining, are point of sales systems in supermarkets. Based on the items bought by the customers, one tries to find items that are frequently bought together. The resulting knowledge could then be used to place these items in the same place, which could positively affect the sales figures. Moreover, one tries to find so-called association rules from frequently reoccurring structures, which are basically implications of the form [L1, ..., Lk] implies [R1, ..., Rm]. Other applications areas other than point of sales systems are Text Mining, Bioinformatics, or Chemoinformatics.