Annotating semantic relations combining facts and opinions

As part of the STATEMENT MAP project, we are constructing a Japanese corpus annotated with the semantic relations bridging facts and opinions that are necessary for online information credibility evaluation. In this paper, we identify the semantic relations essential to this task and discuss how to efficiently collect valid examples from Web documents by splitting complex sentences into fundamental units of meaning called "statements" and annotating relations at the statement level. We present a statement annotation scheme and examine its reliability by annotating around 1,500 pairs of statements. We are preparing the corpus for release this winter.