(s) -> o, each of which recognizes the possible object(s) of a given subject under a specific relation. Under this framework, relational triple extraction is a two-step process: first we identify all ...

Under this framework, relational triple extraction is a two-step process: first we identify all possible subjects in a sentence; then for each subject, we apply relation-specific taggers to ...