Search

The large majority of users accesses Web pages without considering the quality of their contents. In the Web environment, users are not provided tools that measure the quality and reliability of information, while Web information is often non reliable, as it is gathered from different sources that may not be integrated and consistent with each other. In distributed systems, such as the Internet, databases belong to different domains, are built according to different requirements and are updated at times and with procedures that depend on the specific context. Consequently, when information is integrated, problems concerning data consistency, accuracy, usability, and timeliness can arise. This paper addresses these issues and proposes a tool for the evaluation of data quality on the Web and for the management of the related certification process. For this purpose, the paper defines a data quality representation model and a language for the description and implementation of quality metadata by comparing different standards from the literature.

The eighth international workshop on Quality in Database was held in Singapore, on September 13th, 2010 and co-located with the 36th Conference on Very Large DataBase (VLDB). The main objective of the workshop was to address the challenge to detect data anomalies and assess, monitor, improve, and maintain the quality of information. The keynote speech, titled 'SOLOMON: Seeking the Truth Via Copying Detection' was delivered by Xin Luna Dong, AT&T. The author presented the SOLOMON tool that supports the discovery of copying relationships between structured data source to improve data integration features. The goal of 'Deriving Effectiveness Measures for Data Quality Rules' by Lei Jiang, Alex Borgida, Daniele Barone, John Mylopoulos is the evaluation of data quality rules. Peter Yeh, and Colin Puri show in 'Discovering Conditional Functional Dependencies to Detect Data Inconsistencies' an approach that exploits conditional functional dependencies for detecting inconsistencies in data and hence improves data quality.

The paper discusses two concepts that have been associated with various approaches to data and information, namely capacity and value, focusing on data base architectures, and on two types of technologies diffusely used in integration projects, namely data integration, in the area of Enterprise Information Integration, and publish & subscribe, in the area of Enterprise Application Integration. Furthermore, the paper proposes and discusses a unifying model for information capacity and value, that considers also quality constraints and run time costs of the data base architecture.

Measuring and improving data quality in an organization or in a group of interacting organizations is a complex task. Several methodologies have been developed in the past providing a basis for the definition of a complete data quality program applying assessment and improvement techniques in order to guarantee high data quality levels. Since the main limitation of existing approaches is their specialization on specific issues or contexts, this paper presents the Comprehensive Data Quality (CDQ) methodology that aims at integrating and enhancing the phases, techniques and tools proposed by previous approaches. CDQ methodology is conceived to be at the same time complete, flexible and simple to apply. Completeness is achieved by considering existing techniques and tools and integrating them in a framework that can work in both intra and inter organizational contexts, and can be applied to all types of data. The methodology is flexible since it supports the user in the selection of the most suitable techniques and tools within each phase and in any context. Finally, CDQ is simple since It is organized in phases and each phase is char-acterized by a specific goal and techniques to apply. The methodology is explained by means of a running example. © 2006 IEEE.