Detecting Abnormal Data for Ontology Based Information Integration

To better support information integration on Semantic Web data with varying degrees of quality, this paper proposes an approach to detect triples which reflect some sort of er-ror. In particular, erroneous triples may occur due to fac-tual errors in the original data source, misuse of the ontol-ogy by the original data source, or errors in the integra-tion process. Although diagnosing such errors is a difficult problem, we propose that the degree to which a triple de-viates from similar triples can be an important heuristic for identifying errors. We detect such “abnormal triples” by learning probabilistic rules from the reference data and checking to what extent these rules agree with the triples. The system consists of two components for two types of ab-normal relational descriptions that a Semantic Web state-ment could have, whether accidentally or maliciously: a statement could relate two resources that are unlikely to have anything in common or an inappropriate predicate could be used to describe the relation between the two re-sources. The classification technique is adopted to learn statistical characteristics for detecting a suspect resource pair, i.e. there is no significant relation between the subject and the object in the statement. For the suspect usages of a predicate, the system learns semantic patterns for each predicate from indirect semantic connections between the subject / object pairs