arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2403.09259cs.CLcs.LG
收藏

标注还是不标注:用于神经机器翻译的混合主动学习

To Label or Not to Label: Hybrid Active Learning for Neural Machine Translation

Lahore University of Management Sciences(拉合尔管理科学大学)

Abdul Hameed Azeemi, Ihsan Ayyub Qazi, Agha Ali Raza

AI总结:

本文提出混合不确定性与多样性采样(HUDS)策略,用于NMT领域适应,通过结合不确定性和多样性选择句子,显著降低了标注成本并提升了翻译性能。

摘要:

主动学习(AL)技术通过从未标注数据中选择较小的代表性子集进行标注,降低了训练神经机器翻译(NMT)模型的标注成本。多样性采样技术选择异质实例,而不确定性采样方法选择模型不确定性最高的实例。这两种方法都有局限性——多样性方法可能提取多样但平凡的示例,而不确定性采样可能产生重复且无信息量的实例。为了弥合这一差距,我们提出了混合不确定性与多样性采样(HUDS),这是一种用于NMT领域适应的AL策略,它结合了不确定性和多样性进行句子选择。HUDS计算未标注句子的不确定性分数,并随后对其进行分层。然后,它对每个分层内的句子嵌入进行聚类,并通过到质心的距离计算多样性分数。接着,使用结合了不确定性和多样性的加权混合分数来选择每个AL迭代中用于标注的顶级实例。在多领域德语-英语和法语-英语数据集上的实验表明,HUDS的性能优于其他强大的AL基线。我们分析了HUDS的句子选择,并表明它在早期AL迭代中优先选择具有高模型不确定性的多样实例进行标注。

英文摘要:

Active learning (AL) techniques reduce labeling costs for training neural machine translation (NMT) models by selecting smaller representative subsets from unlabeled data for annotation. Diversity sampling techniques select heterogeneous instances, while uncertainty sampling methods select instances with the highest model uncertainty. Both approaches have limitations - diversity methods may extract varied but trivial examples, while uncertainty sampling can yield repetitive, uninformative instances. To bridge this gap, we propose Hybrid Uncertainty and Diversity Sampling (HUDS), an AL strategy for domain adaptation in NMT that combines uncertainty and diversity for sentence selection. HUDS computes uncertainty scores for unlabeled sentences and subsequently stratifies them. It then clusters sentence embeddings within each stratum and computes diversity scores by distance to the centroid. A weighted hybrid score that combines uncertainty and diversity is then used to select the top instances for annotation in each AL iteration. Experiments on multi-domain German-English and French-English datasets demonstrate the better performance of HUDS over other strong AL baselines. We analyze the sentence selection with HUDS and show that it prioritizes diverse instances having high model uncertainty for annotation in early AL iterations.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑