The Universal Dependencies (UD) project seeks to develop cross-linguistically consistent treebank annotation of morphology and syntax for multiple languages. The first version of the dataset was released in 2015 and consisted of 10 treebanks over 10 languages. Version 2.7 released in 2020 consists of 183 treebanks over 104 languages. The annotation consists of UPOS (universal part-of-speech tags), XPOS (language-specific part-of-speech tags), Feats (universal morphological features), Lemmas, dependency heads and universal dependency labels.

505 PAPERS • 12 BENCHMARKS

The MPII Human Pose Dataset for single person pose estimation is composed of about 25K images of which 15K are training samples, 3K are validation samples and 7K are testing samples (which labels are withheld by the authors). The images are taken from YouTube videos covering 410 different human activities and the poses are manually annotated with up to 16 body joints.

463 PAPERS • 4 BENCHMARKS

The Cross-lingual Natural Language Inference (XNLI) corpus is the extension of the Multi-Genre NLI (MultiNLI) corpus to 15 languages. The dataset was created by manually translating the validation and test sets of MultiNLI into each of those 15 languages. The English training set was machine translated for all languages. The dataset is composed of 122k train, 2490 validation and 5010 test examples.

328 PAPERS • 10 BENCHMARKS

Common Voice is an audio dataset that consists of a unique MP3 and corresponding text file. There are 9,283 recorded hours in the dataset. The dataset also includes demographic metadata like age, sex, and accent. The dataset consists of 7,335 validated hours in 60 languages.

314 PAPERS • 164 BENCHMARKS

MuST-C currently represents the largest publicly available multilingual corpus (one-to-many) for speech translation. It covers eight language directions, from English to German, Spanish, French, Italian, Dutch, Portuguese, Romanian and Russian. The corpus consists of audio, transcriptions and translations of English TED talks, and it comes with a predefined training, validation and test split.

194 PAPERS • 2 BENCHMARKS

XQuAD (Cross-lingual Question Answering Dataset) is a benchmark dataset for evaluating cross-lingual question answering performance. The dataset consists of a subset of 240 paragraphs and 1190 question-answer pairs from the development set of SQuAD v1.1 (Rajpurkar et al., 2016) together with their professional translations into ten languages: Spanish, German, Greek, Russian, Turkish, Arabic, Vietnamese, Thai, Chinese, and Hindi. Consequently, the dataset is entirely parallel across 11 languages.

170 PAPERS • 1 BENCHMARK

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.

96 PAPERS • NO BENCHMARKS YET

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.

57 PAPERS • 3 BENCHMARKS

OSCAR or Open Super-large Crawled ALMAnaCH coRpus is a huge multilingual corpus obtained by language classification and filtering of the Common Crawl corpus using the goclassy architecture. The dataset used for training multilingual models such as BART incorporates 138 GB of text.

55 PAPERS • NO BENCHMARKS YET

WikiLingua includes ~770k article and summary pairs in 18 languages from WikiHow. Gold-standard article-summary alignments across languages are extracted by aligning the images that are used to describe each how-to step in an article.

50 PAPERS • 5 BENCHMARKS

XL-Sum is a comprehensive and diverse dataset for abstractive summarization comprising 1 million professionally annotated article-summary pairs from BBC, extracted using a set of carefully designed heuristics. The dataset covers 44 languages ranging from low to high-resource, for many of which no public dataset is currently available. XL-Sum is highly abstractive, concise, and of high quality, as indicated by human and intrinsic evaluation.

41 PAPERS • NO BENCHMARKS YET

A large-scale MultiLingual SUMmarization dataset. Obtained from online newspapers, it contains 1.5M+ article/summary pairs in five different languages -- namely, French, German, Spanish, Russian, Turkish. Together with English newspapers from the popular CNN/Daily mail dataset, the collected data form a large scale multilingual dataset which can enable new research directions for the text summarization community.

40 PAPERS • 5 BENCHMARKS

Multilingual Knowledge Questions and Answers (MKQA) is an open-domain question answering evaluation set comprising 10k question-answer pairs aligned across 26 typologically diverse languages (260k question-answer pairs in total). The goal of this dataset is to provide a challenging benchmark for question answering quality across a wide set of languages. Answers are based on a language-independent data representation, making results comparable across languages and independent of language-specific passages. With 26 languages, this dataset supplies the widest range of languages to-date for evaluating question answering.

37 PAPERS • NO BENCHMARKS YET

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.

32 PAPERS • NO BENCHMARKS YET

News translation is a recurring WMT task. The test set is a collection of parallel corpora consisting of about 1500 English sentences translated into 5 languages (Czech, German, Finnish, Romanian, Russian, Turkish) and additional 1500 sentences from each of the 5 languages translated to English. For Romanian a third of the test set were released as a development set instead. For Turkish additional 500 sentence development set was released. The sentences were selected from dozens of news websites and translated by professional translators. The training data consists of parallel corpora to train translation models, monolingual corpora to train language models and development sets for tuning. Some training corpora were identical from WMT 2015 (Europarl, United Nations, French-English 10⁹ corpus, Common Crawl, Russian-English parallel data provided by Yandex, Wikipedia Headlines provided by CMU) and some were update (CzEng v1.6pre, News Commentary v11, monolingual news data). Additionally,

23 PAPERS • 8 BENCHMARKS

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.

21 PAPERS • 13 BENCHMARKS

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 lan

20 PAPERS • 2 BENCHMARKS

CVSS is a massively multilingual-to-English speech to speech translation (S2ST) corpus, covering sentence-level parallel S2ST pairs from 21 languages into English. CVSS is derived from the Common Voice speech corpus and the CoVoST 2 speech-to-text translation (ST) corpus, by synthesizing the translation text from CoVoST 2 into speech using state-of-the-art TTS systems

18 PAPERS • 1 BENCHMARK

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.

18 PAPERS • 3 BENCHMARKS

Belebele is a multiple-choice machine reading comprehension (MRC) dataset spanning 122 language variants. This dataset enables the evaluation of mono- and multi-lingual models in high-, medium-, and low-resource languages. Each question has four multiple-choice answers and is linked to a short passage from the FLORES-200 dataset. The human annotation procedure was carefully curated to create questions that discriminate between different levels of generalizable language comprehension and is reinforced by extensive quality checks. While all questions directly relate to the passage, the English dataset on its own proves difficult enough to challenge state-of-the-art language models. Being fully parallel, this dataset enables direct comparison of model performance across all languages. Belebele opens up new avenues for evaluating and analyzing the multilingual abilities of language models and NLP systems.

17 PAPERS • NO BENCHMARKS YET

license: apache-2.0 tags: human-feedback size_categories: 100K<n<1M pretty_name: OpenAssistant Conversations

14 PAPERS • NO BENCHMARKS YET

X-FACT is a large publicly available multilingual dataset for factual verification of naturally existing real-world claims. The dataset contains short statements in 25 languages and is labeled for veracity by expert fact-checkers. The dataset includes a multilingual evaluation benchmark that measures both out-of-domain generalization, and zero-shot capabilities of the multilingual models.

13 PAPERS • 1 BENCHMARK

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).

13 PAPERS • NO BENCHMARKS YET

News translation is a recurring WMT task. The test set is a collection of parallel corpora consisting of about 1500 English sentences translated into 5 languages (Chinese, Czech, Estonian, German, Finnish, Russian, Turkish) and additional 1500 sentences from each of the 7 languages translated to English. The sentences were selected from dozens of news websites and translated by professional translators.

8 PAPERS • NO BENCHMARKS YET

RELX is a benchmark dataset for cross-lingual relation classification in English, French, German, Spanish and Turkish.

6 PAPERS • NO BENCHMARKS YET

XL-BEL is a benchmark for cross-lingual biomedical entity linking (XL-BEL). The benchmark spans 10 typologically diverse languages.

6 PAPERS • NO BENCHMARKS YET

The DISRPT 2019 workshop introduces the first iteration of a cross-formalism shared task on discourse unit segmentation. Since all major discourse parsing frameworks imply a segmentation of texts into segments, learning segmentations for and from diverse resources is a promising area for converging methods and insights. We provide training, development and test datasets from all available languages and treebanks in the RST, SDRT and PDTB formalisms, using a uniform format. Because different corpora, languages and frameworks use different guidelines for segmentation, the shared task is meant to promote design of flexible methods for dealing with various guidelines, and help to push forward the discussion of standards for discourse units. For datasets which have treebanks, we will evaluate in two different scenarios: with and without gold syntax, or otherwise using provided automatic parses for comparison.

4 PAPERS • NO BENCHMARKS YET

MuMiN is a misinformation graph dataset containing rich social media data (tweets, replies, users, images, articles, hashtags), spanning 21 million tweets belonging to 26 thousand Twitter threads, each of which have been semantically linked to 13 thousand fact-checked claims across dozens of topics, events and domains, in 41 different languages, spanning more than a decade.

4 PAPERS • 3 BENCHMARKS

In this paper, we present AnlamVer, which is a semantic model evaluation dataset for Turkish designed to evaluate word similarity and word relatedness tasks while discriminating those two relations from each other. Our dataset consists of 500 word-pairs annotated by 12 human subjects, and each pair has two distinct scores for similarity and relatedness. Word-pairs are selected to enable the evaluation of distributional semantic models by multiple attributes of words and word-pair relations such as frequency, morphology, concreteness and relation types (e.g., synonymy, antonymy). Our aim is to provide insights to semantic model researchers by evaluating models in multiple attributes. We balance dataset word-pairs by their frequencies to evaluate the robustness of semantic models concerning out-of-vocabulary and rare words problems, which are caused by the rich derivational and inflectional morphology of the Turkish language. (from the original abstract of the dataset paper)

3 PAPERS • NO BENCHMARKS YET

The DISRPT 2021 shared task, co-located with CODI 2021 at EMNLP, introduces the second iteration of a cross-formalism shared task on discourse unit segmentation and connective detection, as well as the first iteration of a cross-formalism discourse relation classification task.

3 PAPERS • NO BENCHMARKS YET

GeoCoV19 is a large-scale Twitter dataset containing more than 524 million multilingual tweets. The dataset contains around 378K geotagged tweets and 5.4 million tweets with Place information. The annotations include toponyms from the user location field and tweet content and resolve them to geolocations such as country, state, or city level. In this case, 297 million tweets are annotated with geolocation using the user location field and 452 million tweets using tweet content.

3 PAPERS • NO BENCHMARKS YET

AM2iCo is a wide-coverage and carefully designed cross-lingual and multilingual evaluation set. It aims to assess the ability of state-of-the-art representation models to reason over cross-lingual lexical-level concept alignment in context for 14 language pairs.

2 PAPERS • NO BENCHMARKS YET

The fetoscopy placenta dataset is associated with our MICCAI2020 publication titled “Deep Placental Vessel Segmentation for Fetoscopic Mosaicking”. The dataset contains 483 frames with ground-truth vessel segmentation annotations taken from six different in vivo fetoscopic procedure videos. The dataset also includes six unannotated in vivo continuous fetoscopic video clips (950 frames) with predicted vessel segmentation maps obtained from the leave-one-out cross-validation of our method.

2 PAPERS • NO BENCHMARKS YET

The Mobile Turkish Scene Text (MTST 200) dataset consists of 200 indoor and outdoor Turkish scene text images.

2 PAPERS • NO BENCHMARKS YET

The dataset contains training and evaluation data for 12 languages: - Vietnamese - Romanian - Latvian - Czech - Polish - Slovak - Irish - Hungarian - French - Turkish - Spanish - Croatian

2 PAPERS • 12 BENCHMARKS

Natural Language Inference in Turkish (NLI-TR) provides translations of two large English NLI datasets into Turkish and had a team of experts validate their translation quality and fidelity to the original labels.

2 PAPERS • NO BENCHMARKS YET

Bianet is a parallel news corpus in Turkish, Kurdish and English It contains 3,214 Turkish articles with their sentence-aligned Kurdish or English translations from the Bianet online newspaper.

1 PAPER • NO BENCHMARKS YET

DivEMT, the first publicly available post-editing study of Neural Machine Translation (NMT) over a typologically diverse set of target languages. Using a strictly controlled setup, 18 professional translators were instructed to translate or post-edit the same set of English documents into Arabic, Dutch, Italian, Turkish, Ukrainian, and Vietnamese. During the process, their edits, keystrokes, editing times and pauses were recorded, enabling an in-depth, cross-lingual evaluation of NMT quality and post-editing effectiveness. Using this new dataset, we assess the impact of two state-of-the-art NMT systems, Google Translate and the multilingual mBART-50 model, on translation productivity.

1 PAPER • NO BENCHMARKS YET

We introduce GLAMI-1M: the largest multilingual image-text classification dataset and benchmark. The dataset contains images of fashion products with item descriptions, each in 1 of 13 languages. Categorization into 191 classes has high-quality annotations: all 100k images in the test set and 75% of the 1M training set were human-labeled. The paper presents baselines for image-text classification showing that the dataset presents a challenging fine-grained classification problem: The best scoring EmbraceNet model using both visual and textual features achieves 69.7% accuracy. Experiments with a modified Imagen model show the dataset is also suitable for image generation conditioned on text.

1 PAPER • 1 BENCHMARK

We introduce HumanEval-XL, a massively multilingual code generation benchmark specifically crafted to address this deficiency. HumanEval-XL establishes connections between 23 NLs and 12 programming languages (PLs), and comprises of a collection of 22,080 prompts with an average of 8.33 test cases. By ensuring parallel data across multiple NLs and PLs, HumanEval-XL offers a comprehensive evaluation platform for multilingual LLMs, allowing the assessment of the understanding of different NLs. Our work serves as a pioneering step towards filling the void in evaluating NL generalization in the area of multilingual code generation. We make our evaluation code and data publicly available at https://github.com/FloatAI/HumanEval-XL.

1 PAPER • NO BENCHMARKS YET

Describe the Marmara Turkish Coreference Corpus, which is an annotation of the whole METU-Sabanci Turkish Treebank with mentions and coreference chains.

1 PAPER • NO BENCHMARKS YET

Mega-COV is a billion-scale dataset from Twitter for studying COVID-19. The dataset is diverse (covers 234 countries), longitudinal (goes as back as 2007), multilingual (comes in 65 languages), and has a significant number of location-tagged tweets (~32M tweets).

1 PAPER • NO BENCHMARKS YET

MultiTACRED is a multilingual version of the large-scale TAC Relation Extraction Dataset. It covers 12 typologically diverse languages from 9 language families, and was created by the Speech & Language Technology group of DFKI by machine-translating the instances of the original TACRED dataset and automatically projecting their entity annotations. For details of the original TACRED's data collection and annotation process, see the Stanford paper. Translations are syntactically validated by checking the correctness of the XML tag markup. Any translations with an invalid tag structure, e.g. missing or invalid head or tail tag pairs, are discarded (on average, 2.3% of the instances).

1 PAPER • NO BENCHMARKS YET

This dataset contains orthographic samples of words in 19 languages (ar, br, de, en, eno, ent, eo, es, fi, fr, fro, it, ko, nl, pt, ru, sh, tr, zh). Each sample contains two text features: a Word (the textual representation of the word according to its orthography) and a Pronunciation (the highest-surface IPA pronunciation of the word as pronunced in its language).

1 PAPER • NO BENCHMARKS YET

A new high accuracy Turkish morphology dataset.

1 PAPER • NO BENCHMARKS YET

TuGebic is a corpus of recordings of spontaneous speech samples from Turkish-German bilinguals, and the compilation of a corpus called TuGebic. Participants in the study were adult Turkish and German bilinguals living in Germany or Turkey at the time of recording in the first half of the 1990s. The data were manually tokenised and normalised, and all proper names (names of participants and places mentioned in the conversations) were replaced with pseudonyms. Token-level automatic language identification was performed, which made it possible to establish the proportions of words from each language.

1 PAPER • NO BENCHMARKS YET

we have prepared a dataset using publicly available TED Talks transcripts [27] and selected the Turkish corpus. The resulting Turkish punctuation restoration dataset currently consists of 146K sentences and 1.8M tokens. The ratio of the train, validation, and test splits are 0.8, 0.1, and 0.1, respectively. Data files contain two columns. The first column has the tokens separated by white space. The second column includes tags for each token.

1 PAPER • NO BENCHMARKS YET

WEATHub is a dataset containing 24 languages. It contains words organized into groups of (target1, target2, attribute1, attribute2) to measure the association target1:target2 :: attribute1:attribute2. For example target1 can be insects, target2 can be flowers. And we might be trying to measure whether we find insects or flowers pleasant or unpleasant. The measurement of word associations is quantified using the WEAT metric in our paper. It is a metric that calculates an effect size (Cohen's d) and also provides a p-value (to measure statistical significance of the results). In our paper, we use word embeddings from language models to perform these tests and understand biased associations in language models across different languages.

1 PAPER • NO BENCHMARKS YET