Multilingual Open Relation Extraction Using Cross-lingual Projection

Human Language Technologies: The 2015 Annual Conference of the North American Chapter of the ACL, pages 1351–1356, Denver, Colorado, May 31 – June 5, 2015. c⃝2015 Association for Computational Linguistics Multilingual Open Relation Extraction Using Cross-lingual Projection Manaal Faruqui Carnegie Mellon University Pittsburgh, PA 15213 mfaruqui@cs.cmu.edu Shankar Kumar Google Inc. New York, NY 10011 shankarkumar@google.com Abstract Open domain relation extraction systems iden- tify relation and argument phrases in a sen- tence without relying on any underlying schema. However, current state-of-the-art re- lation extraction systems are available only for English because of their heavy reliance on linguistic tools such as part-of-speech tag- gers and dependency parsers. We present a cross-lingual annotation projection method for language independent relation extraction. We evaluate our method on a manually annotated test set and present results on three typolog- ically different languages. We release these manual annotations and extracted relations in ten languages from Wikipedia. 1 Introduction Relation extraction (RE) is the task of assigning a semantic relationship between a pair of arguments. The two major types of RE are closed domain and open domain RE. While closed-domain RE systems (Bunescu and Mooney, 2005; Bunescu, 2007; Mintz et al., 2009; Yao and Van Durme, 2014; Berant and Liang, 2014) consider only a closed set of re- lationships between two arguments, open domain systems (Yates et al., 2007; Carlson et al., 2010; Fader et al., 2011; Mausam et al., 2012) use an ar- bitrary phrase to specify a relationship. In this pa- per, we focus on open-domain RE for multiple lan- guages. Although there are advantages to closed domain RE (Banko and Etzioni, 2008), it is expen- sive to construct a closed set of relation types which would be meaningful across multiple languages. Open RE systems extract patterns from sentences in a given language to identify relations. For learn- ing these patterns, the sentences are analyzed using a part of speech tagger, a dependency parser and pos- sibly a named-entity recognizer. In languages other than English, these tools are either unavailable or not accurate enough to be used. In comparison, it is eas- ier to obtain parallel bilingual corpora which can be used to build machine translation systems (Resnik and Smith, 2003; Smith et al., 2013). In this paper, we present a system that performs RE on a sentence in a source language by first trans- lating the sentence to English, performing RE in En- glish, and finally projecting the relation phrase back to the source language sentence. Our system as- sumes the availability of a machine translation sys- tem from a source language to English and an open RE system in English but no any other analysis tool in the source language. The main contributions of this work are: • A pipeline to develop relation extraction sys- tem for any source language. • Extracted open relations in ten languages based on Wikipedia corpus. • Manual judgements for the projected relations in three languages. We first describe our methodology for language independent cross-lingual projection of extracted re- …

Content truncated to 500 words