Regresar

Learning to get the value of quality from web data

Abstract:

The quality of data used in an information system is highly influenced by the quality of data extracted from the sources that the system uses. This feature is particularly sensitive when the sources provide data coming from the web. These web data are extremely dynamic and heterogeneous and they generally lack from a direct responsible about their quality. This work addresses this problem by presenting a proposal to get the values of quality factors from data coming from the web. One important contribution of this paper is the specification of a generic and flexible Quality Factor Ontology (QF-Ontology) able to model quality factors depending not only on the specific application domain but also on the different types of web sources. Moreover, this paper shows how using SWRL the QF-Ontology is exploited to calculate the metrics associated to each quality factor.