Massively Multilingual Language Models for Cross Lingual Fact Extraction from Low Resource Indian Languages

TL;DR

This paper introduces Cross Lingual Fact Extraction (CLFE) from low-resource Indian languages, proposing an end-to-end generative method that effectively extracts factual triples with a notable F1 score, enhancing multilingual knowledge graph enrichment.

Contribution

It defines the CLFE task and presents a novel generative approach that significantly improves fact extraction from low-resource languages.

Findings

Achieved an overall F1 score of 77.46 on CLFE task.

Demonstrated effectiveness of generative models for cross-lingual fact extraction.

Addresses gap in multilingual information extraction for low-resource languages.

Abstract

Massive knowledge graphs like Wikidata attempt to capture world knowledge about multiple entities. Recent approaches concentrate on automatically enriching these KGs from text. However a lot of information present in the form of natural text in low resource languages is often missed out. Cross Lingual Information Extraction aims at extracting factual information in the form of English triples from low resource Indian Language text. Despite its massive potential, progress made on this task is lagging when compared to Monolingual Information Extraction. In this paper, we propose the task of Cross Lingual Fact Extraction(CLFE) from text and devise an end-to-end generative approach for the same which achieves an overall F1 score of 77.46.