L3Cube-IndicSBERT：一种利用多语言 BERT 学习跨语言句子表示的简单方法
L3Cube-IndicSBERT: A simple approach for learning cross-lingual sentence representations using multilingual BERT
AI总结：
提出通过聚合翻译 NLI/STS 数据并微调普通多语言 BERT 的简单方法，构建并发布面向 10 种印度语言的 L3Cube-IndicSBERT，在跨语言与单语相似度任务上优于多个基线。
摘要：
多语言 Sentence-BERT（SBERT）模型将不同语言映射到共同表示空间，可用于跨语言相似度和挖掘任务。我们提出一种简单而有效的方法，利用合成语料库将普通多语言 BERT 模型转换为多语言句子 BERT 模型。我们只是将低资源目标语言的翻译后 NLI 或 STS 数据集聚合在一起，并对普通多语言 BERT 模型执行类似 SBERT 的微调。我们表明，多语言 BERT 模型是天生的跨语言学习者，这种没有显式跨语言训练的简单基线微调方法会产生出色的跨语言特性。我们在 10 种主要印度语言上展示了方法的有效性，并展示了该方法对非印度语言德语和法语的适用性。利用该方法，我们进一步推出 L3Cube-IndicSBERT，这是首个专门面向印度语言印地语、马拉地语、卡纳达语、泰卢固语、马拉雅拉姆语、泰米尔语、古吉拉特语、奥迪亚语、孟加拉语和旁遮普语的多语言句子表示模型。IndicSBERT 展现出强大的跨语言能力，并在印度语言跨语言和单语句子相似度任务上显著优于 LaBSE、LASER 和 paraphrase-multilingual-mpnet-base-v2 等替代方案。我们还发布了每种语言的单语 SBERT 模型，并表明 IndicSBERT 与其单语对应模型表现相当。这些模型已通过嵌入相似度分数和分类准确率进行评估。
英文摘要：
The multilingual Sentence-BERT (SBERT) models map different languages to common representation space and are useful for cross-language similarity and mining tasks. We propose a simple yet effective approach to convert vanilla multilingual BERT models into multilingual sentence BERT models using synthetic corpus. We simply aggregate translated NLI or STS datasets of the low-resource target languages together and perform SBERT-like fine-tuning of the vanilla multilingual BERT model. We show that multilingual BERT models are inherent cross-lingual learners and this simple baseline fine-tuning approach without explicit cross-lingual training yields exceptional cross-lingual properties. We show the efficacy of our approach on 10 major Indic languages and also show the applicability of our approach to non-Indic languages German and French. Using this approach, we further present L3Cube-IndicSBERT, the first multilingual sentence representation model specifically for Indian languages Hindi, Marathi, Kannada, Telugu, Malayalam, Tamil, Gujarati, Odia, Bengali, and Punjabi. The IndicSBERT exhibits strong cross-lingual capabilities and performs significantly better than alternatives like LaBSE, LASER, and paraphrase-multilingual-mpnet-base-v2 on Indic cross-lingual and monolingual sentence similarity tasks. We also release monolingual SBERT models for each of the languages and show that IndicSBERT performs competitively with its monolingual counterparts. These models have been evaluated using embedding similarity scores and classification accuracy.
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。
