“…In traditional relation extraction, the sources of entities usually come from terms in unstructured documents such as Web pages or structured documents such as relational databases. A wide variety of data sources have been used in relation extraction research, e.g., Web pages (Brin, 1998), corpus (Bunescu & Mooney, 2007), and socially generated Wikipedia articles (Nguyen et al, 2007). The semantic and linguistic sources for exploring relations can be a corpus containing the context of entities, and this context information can serve as the basis of relation assignment.…”