Coreference resolution on RDF Graphs generated from Information Extraction: first results

In our research on the use of information extraction to help populate the semantic web, we have encountered significant obstacles to interoperability. One such obstacle is cross-document coreference resolution. In this paper we describe an effort to improve coreference resolution on RDF graphs generated by text analytics. In addition to driving knowledge-base population, our goal is to demonstrate that successfully combining semantic web and natural language processing technologies can offer advantages over either in isolation, and motivates overcoming the obstacles to interoperability. We present some early results that show improvement of coreference resolution using graph-matching algorithms over RDF.