7 - Advanced Pattern Mining

This chapter discusses the advanced methods of frequent pattern mining, which mines more complex forms of frequent patterns and considers user preferences or constraints to speed up the mining process. Frequent pattern mining has reached far beyond the basics due to substantial research, numerous extensions of the problem scope, and broad application studies. An in-depth coverage of methods for mining many kinds of patterns is included elaborating on: multilevel patterns, multidimensional patterns, patterns in continuous data, rare patterns, negative patterns, constrained frequent patterns, frequent patterns in high-dimensional data, colossal patterns, and compressed and approximate patterns. Other pattern mining themes, including mining sequential and structured patterns and mining patterns from spatiotemporal, multimedia, and stream data, are considered more advanced. Pattern mining is a more general term than frequent pattern mining since the former covers rare and negative patterns as well. However, when there is no ambiguity, the two terms are used interchangeably. In addition to mining for basic frequent itemsets and associations, advanced forms of patterns can be mined such as multilevel associations and multidimensional associations, quantitative association rules, rare patterns, and negative patterns. Users can also mine high-dimensional patterns and compressed or approximate patterns. Frequent pattern mining has many diverse applications, ranging from pattern-based data cleaning to pattern-based classification, clustering, and outlier or exception analysis.