Browse Books

Reviewer: Michael Vassilakopoulos

The quality of data affects all aspects of corporate information management, from data processing to decision-making support. Data accuracy, the focus of this book, is the most important element of data quality. Identifying and handling data quality and accuracy problems gets increasing attention from corporations, as they realize the significant financial implications of these problems. In this book, the author, who is an expert on information systems development, presents his experience on these issues in a systematic way. The book consists of three parts. The first part, comprised of three chapters, introduces the concepts of data quality and accuracy and their impact on information systems, surveys the data quality assurance technology, analyzes the aspects of data quality and accuracy, describes techniques for finding inaccurate values, and analyzes the different sources of inaccuracies. Part 2, also comprised of three chapters, focuses on the data quality assurance process. The first chapter of this part presents the structure and methods of a data quality assurance program. This chapter introduces the fundamental distinction between the inside-out method (that starts from the analysis of data for discovering inaccuracies) and the outside-in method (that starts by seeking for negative impacts at the business level, possibly related to data quality problems). The next chapter focuses on the management of the inaccuracy issues arising from the quality facts discovered, while the last chapter of this part presents the aspects related to the business case of a quality assurance program. The last and largest part of the book is more technical. It consists of seven chapters, and presents the data profiling technology, the core technology of the inside-out approach that has emerged over the last few years. The first chapter is an overview of this technology, and the next six chapters present its steps in more detail. As the author notes, "This part is not an exhaustive treatment of the topic of data profiling," however, it provides a clear view of this technology. It would be more helpful to the interested practitioner if citations to more technical texts appeared in this part. The last chapter summarizes the contribution of the book by commenting on a number of messages that the book aims to disseminate to the audience. Two appendices follow that demonstrate the principles of the data profiling technology. This book is very well structured and written, with a gradual presentation of the various aspects of data quality and accuracy. Although it is biased toward the conceptual view (rather than the technical view), it is one of the most practical books on data profiling I've seen. It can serve as a guide to the data quality and accuracy issues that will receive attention in the years to come. This book is valuable for multiple audiences, depending on their needs. The first part could serve as introductory material on data quality and accuracy for a wide audience interested in information technology. Executives and managers would be more interested in the first and second parts, while system designers and developers, database administrators, computer science students, and especially data quality practitioners can benefit by studying the whole book (especially Part 3). Online Computing Reviews Service