Data analysis and pattern recognition in multiple databases (Q2438807)

In this book, interesting approaches from a number of fields relevant to the actual topic of knowledge discovery in multiple databases are presented. The individual chapters are devoted to Synthesizing different extreme association rules, Clustering items in time-stamped databases induced by stability, Synthesizing global patterns, Clustering local frequency items, Mining patterns of select items, Synthesizing global exceptional patterns, Mining icebergs in different time-stamped data sources, Mining calendar-based periodic patterns in time-stamped data, Measuring influence of an item in time-stamped databases. The latest results are referred to, mainly journal papers and presentations at conferences. Some of the previous results were presented in the monograph [\textit{A. Adhikari} et al., Developing multi-database mining applications. London: Springer (2010; Zbl 1231.68010)]. The present book is designed as an independent text. However, at least two chapters are very similar to their versions in the above monograph. Another common feature of the two books is the relative independency of the chapters. Moreover, each of the chapters has its own list of references, which makes the retrieval difficult. The new four chapters on time-stamped data are based on the research performed by the new co-author Jhimli Adhicari. The book provides a lot of issues to study. In the conclusion, some open problems and further challenges are mentioned: creating a data warehouse for multi-database mining, the role of sampling theory, mining multiple stream data, the need for richer patterns, managing and visualization of large sets of patterns, understandability and the correct interpretation of patterns.