EntityCS: Improving Zero-Shot Cross-lingual Transfer with Entity-Centric Code Switching

TL;DR

This paper introduces EntityCS, a novel entity-centric code-switching method that enhances cross-lingual transfer in multilingual models by leveraging entity alignment from Wikidata and Wikipedia, leading to improved performance on entity-centric tasks.

Contribution

EntityCS is the first approach to focus on entity-level code-switching for better semantic and syntactic preservation, utilizing external knowledge bases for data augmentation.

Findings

10% improvement in Fact Retrieval accuracy

Consistent gains across four entity-centric tasks

Enhanced cross-lingual transfer with entity-aware training

Abstract

Accurate alignment between languages is fundamental for improving cross-lingual pre-trained language models (XLMs). Motivated by the natural phenomenon of code-switching (CS) in multilingual speakers, CS has been used as an effective data augmentation method that offers language alignment at the word- or phrase-level, in contrast to sentence-level via parallel instances. Existing approaches either use dictionaries or parallel sentences with word alignment to generate CS data by randomly switching words in a sentence. However, such methods can be suboptimal as dictionaries disregard semantics, and syntax might become invalid after random word switching. In this work, we propose EntityCS, a method that focuses on Entity-level Code-Switching to capture fine-grained cross-lingual semantics without corrupting syntax. We use Wikidata and English Wikipedia to construct an entity-centric CS corpus by switching entities to their counterparts in other languages. We further propose entity-oriented masking strategies during intermediate model training on the EntityCS corpus for improving entity prediction. Evaluation of the trained models on four entity-centric downstream tasks shows consistent improvements over the baseline with a notable increase of 10% in Fact Retrieval. We release the corpus and models to assist research on code-switching and enriching XLMs with external knowledge.