Controlled Language Generation for Language Learning Items

TL;DR

This paper explores using deep pretrained language models to generate high-quality, controlled English language learning items, focusing on diversity, proficiency levels, and grammatical structures, with positive human evaluation results.

Contribution

It introduces novel control methods for language generation tailored to language learning, enabling diverse and proficiency-specific content creation.

Findings

High grammatical scores in generated items

Enhanced diversity and complexity in outputs

Effective control over proficiency and structure

Abstract

This work aims to employ natural language generation (NLG) to rapidly generate items for English language learning applications: this requires both language models capable of generating fluent, high-quality English, and to control the output of the generation to match the requirements of the relevant items. We experiment with deep pretrained models for this task, developing novel methods for controlling items for factors relevant in language learning: diverse sentences for different proficiency levels and argument structure to test grammar. Human evaluation demonstrates high grammatically scores for all models (3.4 and above out of 4), and higher length (24%) and complexity (9%) over the baseline for the advanced proficiency model. Our results show that we can achieve strong performance while adding additional control to ensure diverse, tailored content for individual users.