The paper "Identifying Relations for Open Information Extraction" by Anthony Fader, Stephen Soderland, and Oren Etzioni addresses the issues of incoherent and uninformative extractions in state-of-the-art Open IE systems. To improve these problems, the authors introduce two simple constraints on binary relations expressed by verbs: a syntactic constraint and a lexical constraint. The syntactic constraint ensures that relation phrases start with a verb, end with a preposition, and are contiguous sequences of words. The lexical constraint ensures that relation phrases appear with a minimal number of distinct argument pairs in a large corpus. These constraints are implemented in the ReVerb Open IE system, which significantly improves the precision and recall of extractions compared to previous systems like TextRUNNER and WOE. The paper also includes a detailed analysis of ReVerb's errors, suggesting directions for future work.The paper "Identifying Relations for Open Information Extraction" by Anthony Fader, Stephen Soderland, and Oren Etzioni addresses the issues of incoherent and uninformative extractions in state-of-the-art Open IE systems. To improve these problems, the authors introduce two simple constraints on binary relations expressed by verbs: a syntactic constraint and a lexical constraint. The syntactic constraint ensures that relation phrases start with a verb, end with a preposition, and are contiguous sequences of words. The lexical constraint ensures that relation phrases appear with a minimal number of distinct argument pairs in a large corpus. These constraints are implemented in the ReVerb Open IE system, which significantly improves the precision and recall of extractions compared to previous systems like TextRUNNER and WOE. The paper also includes a detailed analysis of ReVerb's errors, suggesting directions for future work.