Information Quality Assessment Policies in the Context of Open Information Systems

Newer developments in the areas of Enterprise Data Integration and Knowledge
Management are aiming at using information from a large number of
autonomous data sources inside the company, from partner organizations and
from the Web. Integrating data from independent sources raises questions of
data quality and data trustworthiness. Before information should be used,
its quality has to be assessed according to task-specific criteria. The goal of
the thesis is to develop and evaluate a quality-driven information filtering
framework which supports users in their decision whether to trust or to distrust
information. The main objective of the framework is to support a wide
range of different reputation-, context- and content-based quality assessment
policies. In order to faziliate the user’s understanding of the filtering decisions,
the framework creates explanations why data should be trusted. The
concepts developed in this thesis are implmented as part of the TriQL.P Semantic
Web browser and are evaluated in a financial and a scientific data
integration scenario.