Paper: Open Language Learning for Information Extraction

ACL ID D12-1048
Title Open Language Learning for Information Extraction
Venue Conference on Empirical Methods in Natural Language Processing
Session Main Conference
Year 2012

Open Information Extraction (IE) systems ex- tract relational tuples from text, without re- quiring a pre-specified vocabulary, by iden- tifying relation phrases and associated argu- ments in arbitrary sentences. However, state- of-the-art Open IE systems such as REVERB and WOE share two important weaknesses ? (1) they extract only relations that are medi- ated by verbs, and (2) they ignore context, thus extracting tuples that are not asserted as factual. This paper presents OLLIE, a sub- stantially improved Open IE system that ad- dresses both these limitations. First, OLLIE achieves high yield by extracting relations me- diated by nouns, adjectives, and more. Sec- ond, a context-analysis step increases preci- sion by including contextual information from the sentence in the extractions. ...