On the Hallucination in Simultaneous Machine Translation

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Zhong, Meizhi, Chen, Kehai, Xue, Zhengshan, Liu, Lemao, Yang, Mingming, Zhang, Min
Format: Preprint
Veröffentlicht: 2024
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866910482020958208
author Zhong, Meizhi
Chen, Kehai
Xue, Zhengshan
Liu, Lemao
Yang, Mingming
Zhang, Min
author_facet Zhong, Meizhi
Chen, Kehai
Xue, Zhengshan
Liu, Lemao
Yang, Mingming
Zhang, Min
contents It is widely known that hallucination is a critical issue in Simultaneous Machine Translation (SiMT) due to the absence of source-side information. While many efforts have been made to enhance performance for SiMT, few of them attempt to understand and analyze hallucination in SiMT. Therefore, we conduct a comprehensive analysis of hallucination in SiMT from two perspectives: understanding the distribution of hallucination words and the target-side context usage of them. Intensive experiments demonstrate some valuable findings and particularly show that it is possible to alleviate hallucination by decreasing the over usage of target-side information for SiMT.
format Preprint
id arxiv_https___arxiv_org_abs_2406_07239
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle On the Hallucination in Simultaneous Machine Translation
Zhong, Meizhi
Chen, Kehai
Xue, Zhengshan
Liu, Lemao
Yang, Mingming
Zhang, Min
Computation and Language
It is widely known that hallucination is a critical issue in Simultaneous Machine Translation (SiMT) due to the absence of source-side information. While many efforts have been made to enhance performance for SiMT, few of them attempt to understand and analyze hallucination in SiMT. Therefore, we conduct a comprehensive analysis of hallucination in SiMT from two perspectives: understanding the distribution of hallucination words and the target-side context usage of them. Intensive experiments demonstrate some valuable findings and particularly show that it is possible to alleviate hallucination by decreasing the over usage of target-side information for SiMT.
title On the Hallucination in Simultaneous Machine Translation
topic Computation and Language
url https://arxiv.org/abs/2406.07239