Automatic Web-based relational data imputation

摘要: Data incompleteness is one of the most important data quality problems in enterprise information systems. Most existing imputing techniques just deduce approximate values for incomplete attributes by means some specific rules or mathematical methods. Unfortunately, approximation may be far away from truth. Furthermore, when observed inadequate, they will not work well. The World Wide Web (WWW) has become and widely used source. Several current works have proven that using can augment databases. In this paper, we propose a Web-based relational framework, which tries to automatically retrieve real WWW attributes. try take full advantage relations among different kinds objects based on idea same kind things must with their relatives world. Our proposed consist two automatic query formulation algorithms graph-based candidates extraction model. evaluations are high-quality datasets poor-quality dataset prove effectiveness our approaches.