Multilingual Open Relation Extraction Using Cross-lingual Projection

TL;DR

This paper introduces a cross-lingual projection method for open relation extraction that works across multiple languages without relying on language-specific tools, demonstrated on three diverse languages and extended to 61 languages from Wikipedia.

Contribution

The paper proposes a novel language-independent relation extraction approach using annotation projection, enabling relation extraction in multiple languages without linguistic tools.

Findings

Effective relation extraction in three typologically different languages.

Manual annotations and extracted relations released for 61 languages.

Method outperforms previous approaches limited to English.

Abstract

Open domain relation extraction systems identify relation and argument phrases in a sentence without relying on any underlying schema. However, current state-of-the-art relation extraction systems are available only for English because of their heavy reliance on linguistic tools such as part-of-speech taggers 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 typologically different languages. We release these manual annotations and extracted relations in 61 languages from Wikipedia.