Cross-Lingual Transfer

CoNLL-2003 is a named entity recognition dataset released as a part of CoNLL-2003 shared task: language-independent named entity recognition. The data consists of eight files covering two languages: English and German. For each of the languages there is a training file, a development file, a test file and a large file with unannotated data. The English data was taken from the Reuters Corpus. This corpus consists of Reuters news stories between August 1996 and August 1997. For the training and development set, ten days worth of data were taken from the files representing the end of August 1996. For the test set, the texts were from December 1996. The preprocessed raw data covers the month of September 1996. The text for the German data was taken from the ECI Multilingual Text Corpus. This corpus consists of texts in many languages. The portion of data that was used for this task, was extracted from the German newspaper Frankfurter Rundshau. All three of the training, development and test sets were taken from articles written in one week at the end of August 1992. The raw data were taken from the months of September to December 1992. | English data | Articles | Sentences | Tokens | LOC | MISC | ORG | PER | |-------------------|----------|-----------|---------|------|------|------|------| | Training set | 946 | 14,987 | 203,621 | 7140 | 3438 | 6321 | 6600 | | Development set | 216 | 3,466 | 51,362 | 1837 | 922 | 1341 | 1842 | | Test set | 231 | 3,684 | 46,435 | 1668 | 702 | 1661 | 1617 | Number of articles, sentences, tokens and entities (locations, miscellaneous, organizations, and persons) in English data files. | German data | Articles | Sentences | Tokens | LOC | MISC | ORG | PER | |-------------------|----------|-----------|---------|------|------|------|------| | Training set | 553 | 12,705 | 206,931 | 4363 | 2288 | 2427 | 2773 | | Development set | 201 | 3,068 | 51,444 | 1181 | 1010 | 1241 | 1401 | | Test set | 155 | 3,160 | 51,943 | 1035 | 670 | 773 | 1195 | Number of articles, sentences, tokens and entities (locations, miscellaneous, organizations, and persons) in German data files.

The CoNLL dataset is a widely used resource in the field of natural language processing (NLP). The term “CoNLL” stands for Conference on Natural Language Learning. It originates from a series of shared tasks organized at the Conferences of Natural Language Learning.

PAWS-X contains 23,659 human translated PAWS evaluation pairs and 296,406 machine translated training pairs in six typologically distinct languages: French, Spanish, German, Chinese, Japanese, and Korean. All translated pairs are sourced from examples in PAWS-Wiki. Source: Google Research Image Source: https://arxiv.org/pdf/1908.11828v1.pdf

MLQA (MultiLingual Question Answering) is a benchmark dataset for evaluating cross-lingual question answering performance. MLQA consists of over 5K extractive QA instances (12K in English) in SQuAD format in seven languages - English, Arabic, German, Spanish, Hindi, Vietnamese and Simplified Chinese. MLQA is highly parallel, with QA instances parallel between 4 different languages on average. Source: Facebook Research Image Source: https://github.com/facebookresearch/mlqa

This corpus comprises of monolingual data for 100+ languages and also includes data for romanized languages. This was constructed using the urls and paragraph indices provided by the CC-Net repository by processing January-December 2018 Commoncrawl snapshots. Each file comprises of documents separated by double-newlines and paragraphs within the same document separated by a newline. The data is generated using the open source CC-Net repository. Source: Unsupervised Cross-lingual Representation Learning at Scale

The Cross-lingual Choice of Plausible Alternatives (XCOPA) dataset is a benchmark to evaluate the ability of machine learning models to transfer commonsense reasoning across languages. The dataset is the translation and reannotation of the English COPA (Roemmele et al. 2011) and covers 11 languages from 11 families and several areas around the globe. The dataset is challenging as it requires both the command of world knowledge and the ability to generalise to new languages. Source: https://github.com/cambridgeltl/xcopa

WikiANN, also known as PAN-X, is a multilingual named entity recognition dataset. It consists of Wikipedia articles that have been annotated with LOC (location), PER (person), and ORG (organization) tags in the IOB2 format¹². This dataset serves as a valuable resource for training and evaluating named entity recognition models across various languages. For instance, it includes information about notable individuals, places, and organizations mentioned in Wikipedia articles. Researchers and practitioners can use WikiANN to develop and improve natural language processing systems that identify and classify named entities in text. (1) wikiann · Datasets at Hugging Face. https://huggingface.co/datasets/wikiann. (2) wikiann | TensorFlow Datasets. https://tensorflow.google.cn/datasets/catalog/wikiann. (3) wikiann · Datasets at Hugging Face. https://huggingface.co/datasets/wikiann/viewer/en. (4) WikiAnn Dataset | Papers With Code. https://paperswithcode.com/dataset/wikiann-1.

The shared task of CoNLL-2002 concerns language-independent named entity recognition. The types of named entities include: persons, locations, organizations and names of miscellaneous entities that do not belong to the previous three groups. The participants of the shared task were offered training and test data for at least two languages. Information sources other than the training data might have been used in this shared task. Source: CoNLL 2002 Image Source: https://www.aclweb.org/anthology/W02-2024.pdf

MasakhaNER is a collection of Named Entity Recognition (NER) datasets for 10 different African languages. The languages forming this dataset are: Amharic, Hausa, Igbo, Kinyarwanda, Luganda, Luo, Nigerian-Pidgin, Swahili, Wolof, and Yorùbá.

The Cross-lingual TRansfer Evaluation of Multilingual Encoders (XTREME) benchmark was introduced to encourage more research on multilingual transfer learning,. XTREME covers 40 typologically diverse languages spanning 12 language families and includes 9 tasks that require reasoning about different levels of syntax or semantics. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks, and availability of training data. The languages in XTREME are selected to maximize language diversity, coverage in existing tasks, and availability of training data. Among these are many under-studied languages, such as the Dravidian languages Tamil (spoken in southern India, Sri Lanka, and Singapore), Telugu and Malayalam (spoken mainly in southern India), and the Niger-Congo languages Swahili and Yoruba, spoken in Africa.

Tatoeba is a free collection of example sentences with translations geared towards foreign language learners. It is available in more than 400 languages. Its name comes from the Japanese phrase “tatoeba” (例えば), meaning “for example”. It is written and maintained by a community of volunteers through a model of open collaboration. Individual contributors are known as Tatoebans.

TyDiQA is the gold passage version of the Typologically Diverse Question Answering (TyDiWA) dataset, a benchmark for information-seeking question answering, which covers nine languages. The gold passage version is a simplified version of the primary task, which uses only the gold passage as context and excludes unanswerable questions. It is thus similar to XQuAD and MLQA, while being more challenging as questions have been written without seeing the answers, leading to 3× and 2× less lexical overlap compared to XQuAD and MLQA respectively. Source: XTREME

CoVoST is a large-scale multilingual speech-to-text translation corpus. Its latest 2nd version covers translations from 21 languages into English and from English into 15 languages. It has total 2880 hours of speech and is diversified with 78K speakers and 66 accents. Source: CoVoST 2 and Massively Multilingual Speech-to-Text Translation

The Image-Grounded Language Understanding Evaluation (IGLUE) benchmark brings together—by both aggregating pre-existing datasets and creating new ones—visual question answering, cross-modal retrieval, grounded reasoning, and grounded entailment tasks across 20 diverse languages. The benchmark enables the evaluation of multilingual multimodal models for transfer learning, not only in a zero-shot setting, but also in newly defined few-shot learning setups.

Multicultural Reasoning over Vision and Language (MaRVL) is a dataset based on an ImageNet-style hierarchy representative of many languages and cultures (Indonesian, Mandarin Chinese, Swahili, Tamil, and Turkish). The selection of both concepts and images is entirely driven by native speakers. Afterwards, we elicit statements from native speakers about pairs of images. The task consists in discriminating whether each grounded statement is true or false.

Simplified Chinese dataset for NER in The Third International Chinese Language Processing Bakeoff (2006), provided by Microsoft Research Asia (MSRA).

XGLUE is an evaluation benchmark XGLUE,which is composed of 11 tasks that span 19 languages. For each task, the training data is only available in English. This means that to succeed at XGLUE, a model must have a strong zero-shot cross-lingual transfer capability to learn from the English data of a specific task and transfer what it learned to other languages. Comparing to its concurrent work XTREME, XGLUE has two characteristics: First, it includes cross-lingual NLU and cross-lingual NLG tasks at the same time; Second, besides including 5 existing cross-lingual tasks (i.e. NER, POS, MLQA, PAWS-X and XNLI), XGLUE selects 6 new tasks from Bing scenarios as well, including News Classification (NC), Query-Ad Matching (QADSM), Web Page Ranking (WPR), QA Matching (QAM), Question Generation (QG) and News Title Generation (NTG). Such diversities of languages, tasks and task origin provide a comprehensive benchmark for quantifying the quality of a pre-trained model on cross-lingual natural language understanding and generation.

xSID, a new evaluation benchmark for cross-lingual (X) Slot and Intent Detection in 13 languages from 6 language families, including a very low-resource dialect, covering Arabic (ar), Chinese (zh), Danish (da), Dutch (nl), English (en), German (de), Indonesian (id), Italian (it), Japanese (ja), Kazakh (kk), Serbian (sr), Turkish (tr) and an Austro-Bavarian German dialect, South Tyrolean (de-st).

Japanese-English Subtitle Corpus is a large Japanese-English parallel corpus covering the underrepresented domain of conversational dialogue. It consists of more than 3.2 million examples, making it the largest freely available dataset of its kind. The corpus was assembled by crawling and aligning subtitles found on the web. Source: JESC: Japanese-English Subtitle Corpus

A large multilingual benchmark, XL-WiC, featuring gold standards in 12 new languages from varied language families and with different degrees of resource availability, opening room for evaluation scenarios such as zero-shot cross-lingual transfer. Source: XL-WiC: A Multilingual Benchmark for Evaluating Semantic Contextualization

Danish Dependency Treebank (DaNE) is a named entity annotation for the Danish Universal Dependencies treebank using the CoNLL-2003 annotation scheme. Source: DaNE: A Named Entity Resource for Danish

WikiNEuRal is a high-quality automatically-generated dataset for Multilingual Named Entity Recognition.

Europeana Newspapers consists of four datasets with 100 pages each for the languages Dutch, French, German (including Austrian) as part of the Europeana Newspapers project is expected to contribute to the further development and improvement of named entity recognition systems with a focus on historical content. Source: An Open Corpus for Named Entity Recognition in Historic Newspapers

esXNLI is a bilingual NLI dataset. It comprises 2,490 examples from 5 different genres that were originally annotated in Spanish, and translated into English by professional translators. It serves as a counterpoint to XNLI, which was originally annotated in English and translated into 14 other languages, including Spanish. The dataset was conceived to be used in conjunction with the XNLI development set to analyse the effect of translation in cross-lingual transfer learning. Source: https://github.com/artetxem/esxnli

UNER v1 adds an NER annotation layer to 18 datasets (primarily treebanks from UD) and covers 12 geneologically and ty- pologically diverse languages: Cebuano, Danish, German, English, Croatian, Portuguese, Russian, Slovak, Serbian, Swedish, Tagalog, and Chinese4. Overall, UNER v1 contains nine full datasets with training, development, and test splits over eight languages, three evaluation sets for lower-resource languages (TL and CEB), and a parallel evaluation benchmark spanning six languages.