Browse Books

Reviewer: Vasant B. Kaujalgi

Any information system based on a network of computers has complex design and implementation issues. Performance issues may have to be dealt with by trial and error, because of the lack of tools and techniques to model and predict performance. According to the authors, this book is “an attempt to address the retrieval of information from an environment beyond a centralized relational database system.” Query processing is the most important facility of large, complex, distributed databases. The authors focus on multimedia, object-oriented, deductive, parallel, and distributed databases. They also present important retrieval strategies for fuzzy relational databases, text database systems, image databases, and object-oriented database systems. Chapter 1 provides an overview of relational query processing. It covers important topics, including B-trees and hashing. An approach to evaluating the cost of query processing is formulated. Chapter 2 discusses query processing in object-oriented database systems. The coverage is based on the recommendations of the Object Database Management Group. Chapter 3 describes query processing in distributed relational database management systems, with an emphasis on projection and join operations. Multidatabase systems are defined in chapter 4 as tightly coupled federated database systems in which a global schema is constructed from the schemas of local databases. This chapter also presents query processing in detail. The material on multidatabase architectures and on conversion between relational and object-oriented databases is useful. As databases become very large (with sizes in terabytes), parallel processing becomes more relevant. Various architectures and approaches, including independent, pipelined, and partitioned parallelism, are covered in chapter 5. Chapter 6 discusses the application of fuzzy sets and fuzzy logic to databases. It covers membership functions and the conversion of nested queries into unnested queries. Chapter 7 covers query processing in deductive databases, which consist of a database system and a set of rules. Databases are becoming important in CAD/CAM, geographic information systems, and image retrieval. These databases require multidimensional search; traditional search techniques are inadequate. Chapter 8 explains the use of grid-tree and quad-tree approaches for faster retrieval. Chapters 9 and 10 focus on text retrieval and classification, respectively. Large text databases—used in areas such as libraries, law, and medicine—pose complex problems for retrieval. Chapter 9 discusses measures of effectiveness and methods of retrieval. It is possible to create an order of probability list, which will be presented to the user for certain text documents, based on priorities set by the user. The authors discuss the hardware required for these text retrieval strategies. They also cover the classification of text data and clustering. Finally, chapter 11 looks at databases that store still images and videos. These databases are particularly important on the Internet, especially on the World Wide Web. Image retrieval is based on the entity-relationship model, feature levels, or text. A query language for video retrieval is an important subject of this chapter. An interesting medical application is presented as an example. The book has more than 400 excellent references, and each chapter includes bibliographic notes. The book is for nonexperts, and will be an excellent textbook for a graduate course in database management systems for students of computer science or information sciences.