ETC-NLG: End-to-end Topic-Conditioned Natural Language Generation

TL;DR

ETC-NLG introduces an unsupervised method for topic-conditioned natural language generation that works effectively in low-resource settings by leveraging topic modeling annotations without requiring labeled data.

Contribution

The paper presents a novel unsupervised approach for topic-conditioned NLG that operates over unlabeled data, reducing reliance on large labeled datasets and enabling emergent topic modeling.

Findings

Effective in low-resource Italian setting

Comparable performance in Italian and English

Automatic evaluation of conditioning effectiveness

Abstract

Plug-and-play language models (PPLMs) enable topic-conditioned natural language generation by pairing large pre-trained generators with attribute models used to steer the predicted token distribution towards the selected topic. Despite their computational efficiency, PPLMs require large amounts of labeled texts to effectively balance generation fluency and proper conditioning, making them unsuitable for low-resource settings. We present ETC-NLG, an approach leveraging topic modeling annotations to enable fully-unsupervised End-to-end Topic-Conditioned Natural Language Generation over emergent topics in unlabeled document collections. We first test the effectiveness of our approach in a low-resource setting for Italian, evaluating the conditioning for both topic models and gold annotations. We then perform a comparative evaluation of ETC-NLG for Italian and English using a parallel corpus. Finally, we propose an automatic approach to estimate the effectiveness of conditioning on the generated utterances.