Massively Multilingual Language Models for Cross Lingual Fact Extraction from Low Resource Indian Languages
This addresses the gap in knowledge graph enrichment for low-resource languages, which is an incremental improvement over existing monolingual methods.
The paper tackles the problem of extracting factual information from low-resource Indian language texts into English triples, proposing a cross-lingual fact extraction task and achieving an overall F1 score of 77.46 with an end-to-end generative approach.
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.