KNOWLEDGE DISCOVERY IN MULTIPLE DATABASES

Many organizations have an urgent need of mining their multiple databases inherently distributed in branches (distributed data). In particular, as the Web is rapidly becoming an information flood, individuals and organizations can take into account low-cost information and knowledge on the Internet when making decisions. How to efficiently identify quality knowledge from different data sources has become a significant challenge. This challenge has attracted a great many researchers including the au­ thors who have developed a local pattern analysis, a new strategy for dis­ covering some kinds of potentially useful patterns that cannot be mined in traditional multi-database mining techniques. Local pattern analysis deliv­ ers high-performance pattern discovery from multiple databases. There has been considerable progress made on multi-database mining in such areas as hierarchical meta-learning, collective mining, database classification, and pe­ culiarity discovery. While these techniques continue to be future topics of interest concerning multi-database mining, this book focuses on these inter­ esting issues under the framework of local pattern analysis. The book is intended for researchers and students in data mining, dis­ tributed data analysis, machine learning, and anyone else who is interested in multi-database mining. It is also appropriate for use as a text supplement for broader courses that might also involve knowledge discovery in databases and data mining.

From the reviews:

"The book contains the latest on research in database multi-mining (32 papers published after 2000) and offers for consideration a local-pattern analysis framework for pattern discovery from multiple data sources. Starting from the local pattern in multiple data bases, the authors propose ... a new pattern named ‘high-vote’ pattern based on statistical analysis of vote ratio received by a pattern from each branch of the company." (Silviu Craciunas, Zentralblatt MATH, Vol. 1067, 2005)