Detecting and Mitigating Hallucinations in Multilingual Summarisation
Submission Type: Regular Long Paper
Submission Track: Summarization
Submission Track 2: Multilinguality and Linguistic Diversity
Keywords: Summarisation, Multilingual NLP, Hallucination, Natural Language Generation, Faithfulness Evaluation
TL;DR: We develop mFACT to assess the faithfulness of non-English summaries. We show that cross-lingual transfer can boost summarisation performance but amplify hallucinations. We propose a method to reduce hallucinations for cross-lingual transfer.
Abstract: Hallucinations pose a significant challenge to the reliability of neural models for abstractive summarisation. While automatically generated summaries may be fluent, they often lack faithfulness to the original document. This issue becomes even more pronounced in low-resource languages, where summarisation requires cross-lingual transfer. With the existing faithful metrics focusing on English, even measuring the extent of this phenomenon in cross-lingual settings is hard. To address this, we first develop a novel metric, mFACT, evaluating the faithfulness of non-English summaries, leveraging translation-based transfer from multiple English faithfulness metrics. Through extensive experiments in multiple languages, we demonstrate that mFACT is best suited to detect hallucinations compared to alternative metrics. With mFACT, we assess a broad range of multilingual large language models, and find that they all tend to hallucinate often in languages different from English. We then propose a simple but effective method to reduce hallucinations in cross-lingual transfer, which weighs the loss of each training example by its faithfulness score. This method drastically increases both performance and faithfulness according to both automatic and human evaluation when compared to strong baselines for cross-lingual transfer such as MAD-X. Our code and dataset are available at https://github.com/yfqiu-nlp/mfact-summ.
Submission Number: 3617
Loading