Pre-trained multilingual language models (e.g., mBERT, XLM-RoBERTa) have significantly advanced the state-of-the-art for zero-shot cross-lingual information extraction. These language models ubiquitously rely on word segmentation techniques that break a word into smaller constituent subwords. Therefore, all word labeling tasks (e.g. named entity recognition, event detection, etc.), necessitate a pooling strategy that takes the subword representations as input and outputs a representation for the entire word. Taking the task of cross-lingual event detection as a motivating example, we show that the choice of pooling strategy can have a significant impact on the target language performance. For example, the performance varies by up to 16 absolute $f_{1}$ points depending on the pooling strategy when training in English and testing in Arabic on the ACE task. We carry out our analysis with five different pooling strategies across nine languages in diverse multi-lingual datasets. Across configurations, we find that the canonical strategy of taking just the first subword to represent the entire word is usually sub-optimal. On the other hand, we show that attention pooling is robust to language and dataset variations by being either the best or close to the optimal strategy. For reproducibility, we make our code available at https://github.com/isi-boston/ed-pooling.
University of Maryland（马里兰大学）；University of Colorado（科罗拉多大学）；Google Research（谷歌研究院）
Text classification must sometimes be applied in a low-resource language with no labeled training data. However, training data may be available in a related language. We investigate whether character-level knowledge transfer from a related language helps text classification. We present a cross-lingual document classification framework (CACO) that exploits cross-lingual subword similarity by jointly training a character-based embedder and a word-based classifier. The embedder derives vector representations for input words from their written forms, and the classifier makes predictions based on t…
This study addresses unsupervised subword modeling, i.e., learning feature representations that can distinguish subword units of a language. The proposed approach adopts a two-stage bottleneck feature (BNF) learning framework, consisting of autoregressive predictive coding (APC) as a front-end and a DNN-BNF model as a back-end. APC pretrained features are set as input features to a DNN-BNF model. A language-mismatched ASR system is used to provide cross-lingual phone labels for DNN-BNF model training. Finally, BNFs are extracted as the subword-discriminative feature representation. A second ai…
HEIG-VD（沃州应用科学与艺术大学）；EPFL School of Engineering（洛桑联邦理工学院工程学院）
State-of-the-art multilingual systems rely on shared vocabularies that sufficiently cover all considered languages. To this end, a simple and frequently used approach makes use of subword vocabularies constructed jointly over several languages. We hypothesize that such vocabularies are suboptimal due to false positives (identical subwords with different meanings across languages) and false negatives (different subwords with similar meanings). To address these issues, we propose Subword Mapping and Anchoring across Languages (SMALA), a method to construct bilingual subword vocabularies. SMALA e…
This study addresses unsupervised subword modeling, i.e., learning acoustic feature representations that can distinguish between subword units of a language. We propose a two-stage learning framework that combines self-supervised learning and cross-lingual knowledge transfer. The framework consists of autoregressive predictive coding (APC) as the front-end and a cross-lingual deep neural network (DNN) as the back-end. Experiments on the ABX subword discriminability task conducted with the Libri-light and ZeroSpeech 2017 databases showed that our approach is competitive or superior to state-of-…
Reverse dictionary is the task to find the proper target word given the word description. In this paper, we tried to incorporate BERT into this task. However, since BERT is based on the byte-pair-encoding (BPE) subword encoding, it is nontrivial to make BERT generate a word given the description. We propose a simple but effective method to make BERT generate the target word for this specific task. Besides, the cross-lingual reverse dictionary is the task to find the proper target word described in another language. Previous models have to keep two different word embeddings and learn to align t…
This research addresses the problem of acoustic modeling of low-resource languages for which transcribed training data is absent. The goal is to learn robust frame-level feature representations that can be used to identify and distinguish subword-level speech units. The proposed feature representations comprise various types of multilingual bottleneck features (BNFs) that are obtained via multi-task learning of deep neural networks (MTL-DNN). One of the key problems is how to acquire high-quality frame labels for untranscribed training data to facilitate supervised DNN training. It is shown th…
National Institute of Information and Communications Technology（信息通信技术研究所）
Multilingual pretrained models are effective for machine translation and cross-lingual processing because they contain multiple languages in one model. However, they are pretrained after their tokenizers are fixed; therefore it is difficult to change the vocabulary after pretraining. When we extend the pretrained models to new languages, we must modify the tokenizers simultaneously. In this paper, we add new subwords to the SentencePiece tokenizer to apply a multilingual pretrained model to new languages (Inuktitut in this paper). In our experiments, we segmented Inuktitut sentences into subwo…
Multilingual pretrained representations generally rely on subword segmentation algorithms to create a shared multilingual vocabulary. However, standard heuristic algorithms often lead to sub-optimal segmentation, especially for languages with limited amounts of data. In this paper, we take two major steps towards alleviating this problem. First, we demonstrate empirically that applying existing subword regularization methods(Kudo, 2018; Provilkov et al., 2020) during fine-tuning of pre-trained multilingual representations improves the effectiveness of cross-lingual transfer. Second, to take fu…
Harbin Institute of Technology（哈尔滨工业大学）；Microsoft Corporation（微软公司）；Microsoft Research（微软研究院）
Fine-tuning pre-trained cross-lingual language models can transfer task-specific supervision from one language to the others. In this work, we propose to improve cross-lingual fine-tuning with consistency regularization. Specifically, we use example consistency regularization to penalize the prediction sensitivity to four types of data augmentations, i.e., subword sampling, Gaussian noise, code-switch substitution, and machine translation. In addition, we employ model consistency to regularize the models trained with two augmented versions of the same training set. Experimental results on the …
Westlake University（西湖大学）；Westlake Institute for Advanced Study（西湖高等研究院）；Harbin Institute of Technology（哈尔滨工业大学）
Multi-lingual contextualized embeddings, such as multilingual-BERT (mBERT), have shown success in a variety of zero-shot cross-lingual tasks. However, these models are limited by having inconsistent contextualized representations of subwords across different languages. Existing work addresses this issue by bilingual projection and fine-tuning technique. We propose a data augmentation framework to generate multi-lingual code-switching data to fine-tune mBERT, which encourages model to align representations from source and multiple target languages once by mixing their context information. Compa…
HiTZ Center, University of the Basque Country (UPV/EHU)（巴斯克大学 HiTZ 中心）；DeepMind
State-of-the-art unsupervised multilingual models (e.g., multilingual BERT) have been shown to generalize in a zero-shot cross-lingual setting. This generalization ability has been attributed to the use of a shared subword vocabulary and joint training across multiple languages giving rise to deep multilingual abstractions. We evaluate this hypothesis by designing an alternative approach that transfers a monolingual model to new languages at the lexical level. More concretely, we first train a transformer-based masked language model on one language, and transfer it to a new language by learnin…
Most existing machine translation systems operate at the level of words, relying on explicit segmentation to extract tokens. We introduce a neural machine translation (NMT) model that maps a source character sequence to a target character sequence without any segmentation. We employ a character-level convolutional network with max-pooling at the encoder to reduce the length of source representation, allowing the model to be trained at a speed comparable to subword-level models while capturing local regularities. Our character-to-character model outperforms a recently proposed baseline with a s…
Beyond Literal Token Overlap: Token Alignability for Multilinguality
Centre for Information and Language Processing, LMU Munich（慕尼黑大学信息与语言处理中心）；Munich Center for Machine Learning（慕尼黑机器学习中心）；Faculty of Mathematics and Physics, Charles University（查理大学数学与物理学院）；Technical University of Munich（慕尼黑工业大学）
In zero-shot cross-lingual transfer, a supervised NLP task trained on a corpus in one language is directly applicable to another language without any additional training. A source of cross-lingual transfer can be as straightforward as lexical overlap between languages (e.g., use of the same scripts, shared subwords) that naturally forces text embeddings to occupy a similar representation space. Recently introduced cross-lingual language model (XLM) pretraining brings out neural parameter sharing in Transformer-style networks as the most important factor for the transfer. In this paper, we aim …
Borealis AI；Institute for Computer Science and Control（计算机科学与控制研究所）
Contextual word-representations became a standard in modern natural language processing systems. These models use subword tokenization to handle large vocabularies and unknown words. Word-level usage of such systems requires a way of pooling multiple subwords that correspond to a single word. In this paper we investigate how the choice of subword pooling affects the downstream performance on three tasks: morphological probing, POS tagging and NER, in 9 typologically diverse languages. We compare these in two massively multilingual models, mBERT and XLM-RoBERTa. For morphological tasks, the wid…
Idiap Research Institute（Idiap研究所）；University of Edinburgh（爱丁堡大学）；Stellenbosch University（斯泰伦博斯大学）
Subword modeling for zero-resource languages aims to learn low-level representations of speech audio without using transcriptions or other resources from the target language (such as text corpora or pronunciation dictionaries). A good representation should capture phonetic content and abstract away from other types of variability, such as speaker differences and channel noise. Previous work in this area has primarily focused unsupervised learning from target language data only, and has been evaluated only intrinsically. Here we directly compare multiple methods, including some that use only ta…
Successful methods for unsupervised neural machine translation (UNMT) employ crosslingual pretraining via self-supervision, often in the form of a masked language modeling or a sequence generation task, which requires the model to align the lexical- and high-level representations of the two languages. While cross-lingual pretraining works for similar languages with abundant corpora, it performs poorly in low-resource and distant languages. Previous research has shown that this is because the representations are not sufficiently aligned. In this paper, we enhance the bilingual masked language m…
State-of-the-art multilingual models depend on vocabularies that cover all of the languages the model will expect to see at inference time, but the standard methods for generating those vocabularies are not ideal for massively multilingual applications. In this work, we introduce a novel procedure for multilingual vocabulary generation that combines the separately trained vocabularies of several automatically derived language clusters, thus balancing the trade-off between cross-lingual subword sharing and language-specific vocabularies. Our experiments show improvements across languages on key…
Pralekha: Cross-Lingual Document Alignment for Indic Languages
Nilekani Centre at AI4Bharat（AI4Bharat的Nilekani中心）；Microsoft（微软）；National Institute of Information and Communications Technology, Japan（日本信息与通信技术国家研究所）；Indian Institute of Technology, Madras（印度理工学院马德拉斯分校）；Indian Institute of Technology, Bombay（印度理工学院孟买分校）
Large pretrained language models (LMs) have become the central building block of many NLP applications. Training these models requires ever more computational resources and most of the existing models are trained on English text only. It is exceedingly expensive to train these models in other languages. To alleviate this problem, we introduce a novel method -- called WECHSEL -- to efficiently and effectively transfer pretrained LMs to new languages. WECHSEL can be applied to any model which uses subword-based tokenization and learns an embedding for each subword. The tokenizer of the source mo…
(Short version of Abstract) This thesis describes an investigation on unsupervised acoustic modeling (UAM) for automatic speech recognition (ASR) in the zero-resource scenario, where only untranscribed speech data is assumed to be available. UAM is not only important in addressing the general problem of data scarcity in ASR technology development but also essential to many non-mainstream applications, for examples, language protection, language acquisition and pathological speech assessment. The present study is focused on two research problems. The first problem concerns unsupervised discover…
University of Bucharest（布加勒斯特大学）；University Politehnica of Bucharest（布加勒斯特理工大学）
Current benchmark tasks for natural language processing contain text that is qualitatively different from the text used in informal day to day digital communication. This discrepancy has led to severe performance degradation of state-of-the-art NLP models when fine-tuned on real-world data. One way to resolve this issue is through lexical normalization, which is the process of transforming non-standard text, usually from social media, into a more standardized form. In this work, we propose a sentence-level sequence-to-sequence model based on mBART, which frames the problem as a machine transla…
T-FREE: Subword Tokenizer-Free Generative LLMs via Sparse Representations for Memory-Efficient Embeddings
Aleph Alpha；Technical University Darmstadt（达姆施塔特工业大学）；Hessian Center for Artificial Intelligence (hessian.AI)（黑森人工智能中心）；German Research Center for Artificial Intelligence (DFKI)（德国人工智能研究中心）
School of Computing and Information Systems, The University of Melbourne（墨尔本大学计算与信息系统学院）；NTT Communication Science Laboratories（NTT通信科学实验室）；RIKEN Center for Advanced Intelligence Project (AIP)（理化学研究所革新智能综合研究中心）
We propose a new approach for learning contextualised cross-lingual word embeddings based on a small parallel corpus (e.g. a few hundred sentence pairs). Our method obtains word embeddings via an LSTM encoder-decoder model that simultaneously translates and reconstructs an input sentence. Through sharing model parameters among different languages, our model jointly trains the word embeddings in a common cross-lingual space. We also propose to combine word and subword embeddings to make use of orthographic similarities across different languages. We base our experiments on real-world data from …
In natural language processing, the deep learning revolution has shifted the focus from conventional hand-crafted symbolic representations to dense inputs, which are adequate representations learned automatically from corpora. However, particularly when working with low-resource languages, small amounts of symbolic lexical resources such as user-generated lexicons are often available even when gold-standard corpora are not. Such additional linguistic information is though often neglected, and recent neural approaches to cross-lingual tagging typically rely only on word and subword embeddings. …
Low-resource languages, such as Baltic languages, benefit from Large Multilingual Models (LMs) that possess remarkable cross-lingual transfer performance capabilities. This work is an interpretation and analysis study into cross-lingual representations of Multilingual LMs. Previous works hypothesized that these LMs internally project representations of different languages into a shared cross-lingual space. However, the literature produced contradictory results. In this paper, we revisit the prior work claiming that "BERT is not an Interlingua" and show that different languages do converge to a…
Center for Artificial Intelligence Research (CAiRE)（人工智能研究中心（CAiRE））；The Hong Kong University of Science and Technology（香港科技大学）
In countries that speak multiple main languages, mixing up different languages within a conversation is commonly called code-switching. Previous works addressing this challenge mainly focused on word-level aspects such as word embeddings. However, in many cases, languages share common subwords, especially for closely related languages, but also for languages that are seemingly irrelevant. Therefore, we propose Hierarchical Meta-Embeddings (HME) that learn to combine multiple monolingual word-level and subword-level embeddings to create language-agnostic lexical representations. On the task of …
School of Information Science and Technology, Vidyasirimedhi Institution of Science and Technology（威集玛迪科学技术大学信息科学技术学院）；PyThaiNLP
Transformer-based language models, more specifically BERT-based architectures have achieved state-of-the-art performance in many downstream tasks. However, for a relatively low-resource language such as Thai, the choices of models are limited to training a BERT-based model based on a much smaller dataset or finetuning multi-lingual models, both of which yield suboptimal downstream performance. Moreover, large-scale multi-lingual pretraining does not take into account language-specific features for Thai. To overcome these limitations, we pretrain a language model based on RoBERTa-base architect…
