| Data Mining, or knowledge discovery in databases, has been popularly recognized as an important research field with broad potential applications. To have an idea about the issues and plausible techniques for this new area, one may view data mining through four perspectives, namely the knowledge to be mined, the available tools, the types of databases to deal with, and applications.;This report provides an integrated overview on the data mining techniques developed recently. The kinds of knowledge to be mined, including association rules, characteristic rules, classification rules, discriminant rules, clustering rules, evolution rules and deviation rules, are reviewed in detail along with challenging issues and emerging algorithms. The available tools that are borrowed from several disciplines, such as databases, statistics, pattern recognition and machine learning of Artificial Intelligence (AI), and visualization, are examined as a hierarchical tree. Different types of databases, including relational databases, object-oriented databases, inductive databases, spatial databases, temporal databases and scientific databases, internet information systems, etc. are summarized over their data structures, operations, data mining issues, strategies and challenges. Several other related issues on data cleaning, complexity, and privacy are discussed. Three successful data mining examples are provided along with other reported systems and applications. The data mining system architectures are outlined, following which the data mining future research directions and open problems are pointed out and some final thoughts are given. |
