Reconciling Information in DBpedia through a Question Answering System
Abstract
Results obtained querying language-specic DBpedia chap-
ters SPARQL endpoints for the same query can be related by several het-
erogenous relations, or contain an inconsistent set of information about
the same topic. To overcome this issue in question answering systems
over language-specic DBpedia chapters, we propose the RADAR frame-
work for information reconciliation. Starting from a categorization of
the possible relations among the resulting instances, such framework: (i)
classies such relations, (ii) reconciles the obtained information using
argumentation theory, (iii) ranks the alternative results depending on
the condence of the source in case of inconsistencies, and (iv) explains
the reasons underlying the proposed ranking.