{ "name": "TaTA", "summary": "Existing data-to-text generation datasets are mostly limited to English. Table-to-Text in African languages (TaTA) addresses this lack of data as the first large multilingual table-to-text dataset with a focus on African languages. TaTA was created by transcribing figures and accompanying text in bilingual reports by the Demographic and Health Surveys Program, followed by professional translation to make the dataset fully parallel. TaTA includes 8,700 examples in nine languages including four African languages (Hausa, Igbo, Swahili, and Yor\u00f9b\u00e1) and a zero-shot test language (Russian). \n\nYou can load the dataset via:\n```\nimport datasets\ndata = datasets.load_dataset('GEM/TaTA')\n```\nThe data loader can be found [here](https://huggingface.co/datasets/GEM/TaTA).", "sections": [ { "title": "Dataset Overview", "level": 2, "subsections": [ { "title": "Where to find the Data and its Documentation", "level": 3, "fields": [ { "title": "Webpage", "level": 4, "content": "[Github](https://github.com/google-research/url-nlp)", "flags": [], "info": "What is the webpage for the dataset (if it exists)?", "scope": "telescope" }, { "title": "Download", "level": 4, "content": "[Github](https://github.com/google-research/url-nlp)", "flags": [ "" ], "info": "What is the link to where the original dataset is hosted?", "scope": "telescope" }, { "title": "Paper", "level": 4, "content": "[ArXiv](https://arxiv.org/abs/2211.00142)", "flags": [ "" ], "info": "What is the link to the paper describing the dataset (open access preferred)?", "scope": "telescope" }, { "title": "BibTex", "level": 4, "content": "```\n@misc{gehrmann2022TaTA,\n Author = {Sebastian Gehrmann and Sebastian Ruder and Vitaly Nikolaev and Jan A. Botha and Michael Chavinda and Ankur Parikh and Clara Rivera},\n Title = {TaTa: A Multilingual Table-to-Text Dataset for African Languages},\n Year = {2022},\n Eprint = {arXiv:2211.00142},\n}\n```", "flags": [ "" ], "info": "Provide the BibTex-formatted reference for the dataset. Please use the correct published version (ACL anthology, etc.) instead of google scholar created Bibtex.", "scope": "microscope" }, { "title": "Contact Name", "level": 4, "content": "Sebastian Ruder", "flags": [ "quick" ], "info": "If known, provide the name of at least one person the reader can contact for questions about the dataset.", "scope": "periscope" }, { "title": "Contact Email", "level": 4, "content": "ruder@google.com", "flags": [ "" ], "info": "If known, provide the email of at least one person the reader can contact for questions about the dataset.", "scope": "periscope" }, { "title": "Has a Leaderboard?", "level": 4, "content": "yes", "flags": [ "" ], "info": "Does the dataset have an active leaderboard?", "scope": "telescope" }, { "title": "Leaderboard Link", "level": 4, "content": "[Github](https://github.com/google-research/url-nlp)", "flags": [ "" ], "info": "Provide a link to the leaderboard.", "scope": "periscope" }, { "title": "Leaderboard Details", "level": 4, "content": "The paper introduces a metric StATA which is trained on human ratings and which is used to rank approaches submitted to the leaderboard.", "flags": [ "" ], "info": "Briefly describe how the leaderboard evaluates models.", "scope": "microscope" } ] }, { "title": "Languages and Intended Use", "level": 3, "fields": [ { "title": "Multilingual?", "level": 4, "content": "yes", "flags": [ "quick" ], "info": "Is the dataset multilingual?", "scope": "telescope" }, { "title": "Covered Languages", "level": 4, "content": "`English`, `Portuguese`, `Arabic`, `French`, `Hausa`, `Swahili (macrolanguage)`, `Igbo`, `Yoruba`, `Russian`", "flags": [ "quick" ], "info": "What languages/dialects are covered in the dataset?", "scope": "telescope" }, { "title": "Whose Language?", "level": 4, "content": "The language is taken from reports by the demographic and health surveys program.", "flags": [ "" ], "info": "Whose language is in the dataset?", "scope": "periscope" }, { "title": "License", "level": 4, "content": "cc-by-sa-4.0: Creative Commons Attribution Share Alike 4.0 International", "flags": [ "quick" ], "info": "What is the license of the dataset?", "scope": "telescope" }, { "title": "Intended Use", "level": 4, "content": "The dataset poses significant reasoning challenges and is thus meant as a way to asses the verbalization and reasoning capabilities of structure-to-text models.", "flags": [ "" ], "info": "What is the intended use of the dataset?", "scope": "microscope" }, { "title": "Primary Task", "level": 4, "content": "Data-to-Text", "flags": [ "" ], "info": "What primary task does the dataset support?", "scope": "telescope" }, { "title": "Communicative Goal", "level": 4, "content": "Summarize key information from a table in a single sentence.\n", "flags": [ "quick" ], "info": "Provide a short description of the communicative goal of a model trained for this task on this dataset.", "scope": "periscope" } ] }, { "title": "Credit", "level": 3, "fields": [ { "title": "Curation Organization Type(s)", "level": 4, "content": "`industry`", "flags": [ "" ], "info": "In what kind of organization did the dataset curation happen?", "scope": "telescope" }, { "title": "Curation Organization(s)", "level": 4, "content": "Google Research", "flags": [ "" ], "info": "Name the organization(s).", "scope": "periscope" }, { "title": "Dataset Creators", "level": 4, "content": "Sebastian Gehrmann, Sebastian Ruder , Vitaly Nikolaev, Jan A. Botha, Michael Chavinda, Ankur Parikh, Clara Rivera", "flags": [ "" ], "info": "Who created the original dataset? List the people involved in collecting the dataset and their affiliation(s).", "scope": "microscope" }, { "title": "Funding", "level": 4, "content": "Google Research", "flags": [ "" ], "info": "Who funded the data creation?", "scope": "microscope" }, { "title": "Who added the Dataset to GEM?", "level": 4, "content": "Sebastian Gehrmann (Google Research)", "flags": [ "" ], "info": "Who contributed to the data card and adding the dataset to GEM? List the people+affiliations involved in creating this data card and who helped integrate this dataset into GEM.", "scope": "microscope" } ] }, { "title": "Dataset Structure", "level": 3, "fields": [ { "title": "Data Fields", "level": 4, "content": "- `example_id`: The ID of the example. Each ID (e.g., `AB20-ar-1`) consists of three parts: the document ID, the language ISO 639-1 code, and the index of the table within the document.\n- `title`: The title of the table.\n- `unit_of_measure`: A description of the numerical value of the data. E.g., percentage of households with clean water.\n- `chart_type`: The kind of chart associated with the data. We consider the following (normalized) types: horizontal bar chart, map chart, pie graph, tables, line chart, pie chart, vertical chart type, line graph, vertical bar chart, and other.\n- `was_translated`: Whether the table was transcribed in the original language of the report or translated.\n- `table_data`: The table content is a JSON-encoded string of a two-dimensional list, organized by row, from left to right, starting from the top of the table. Number of items varies per table. Empty cells are given as empty string values in the corresponding table cell.\n- `table_text`: The sentences forming the description of each table are encoded as a JSON object. In the case of more than one sentence, these are separated by commas. Number of items varies per table.\n- `linearized_input`: A single string that contains the table content separated by vertical bars, i.e., |. Including title, unit of measurement, and the content of each cell including row and column headers in between brackets, i.e., (Medium Empowerment, Mali, 17.9).", "flags": [ "" ], "info": "List and describe the fields present in the dataset.", "scope": "telescope" }, { "title": "Reason for Structure", "level": 4, "content": "The structure includes all available information for the infographics on which the dataset is based.", "flags": [], "info": "How was the dataset structure determined?", "scope": "microscope" }, { "title": "How were labels chosen?", "level": 4, "content": "Annotators looked through English text to identify sentences that describe an infographic. They then identified the corresponding location of the parallel non-English document. All sentences were extracted.", "flags": [ "" ], "info": "How were the labels chosen?", "scope": "microscope" }, { "title": "Example Instance", "level": 4, "content": "```\n{\n \"example_id\": \"FR346-en-39\",\n \"title\": \"Trends in early childhood mortality rates\",\n \"unit_of_measure\": \"Deaths per 1,000 live births for the 5-year period before the survey\",\n \"chart_type\": \"Line chart\",\n \"was_translated\": \"False\",\n \"table_data\": \"[[\\\"\\\", \\\"Child mortality\\\", \\\"Neonatal mortality\\\", \\\"Infant mortality\\\", \\\"Under-5 mortality\\\"], [\\\"1990 JPFHS\\\", 5, 21, 34, 39], [\\\"1997 JPFHS\\\", 6, 19, 29, 34], [\\\"2002 JPFHS\\\", 5, 16, 22, 27], [\\\"2007 JPFHS\\\", 2, 14, 19, 21], [\\\"2009 JPFHS\\\", 5, 15, 23, 28], [\\\"2012 JPFHS\\\", 4, 14, 17, 21], [\\\"2017-18 JPFHS\\\", 3, 11, 17, 19]]\",\n \"table_text\": [\n \"neonatal, infant, child, and under-5 mortality rates for the 5 years preceding each of seven JPFHS surveys (1990 to 2017-18).\",\n \"Under-5 mortality declined by half over the period, from 39 to 19 deaths per 1,000 live births.\",\n \"The decline in mortality was much greater between the 1990 and 2007 surveys than in the most recent period.\",\n \"Between 2012 and 2017-18, under-5 mortality decreased only modestly, from 21 to 19 deaths per 1,000 live births, and infant mortality remained stable at 17 deaths per 1,000 births.\"\n ],\n \"linearized_input\": \"Trends in early childhood mortality rates | Deaths per 1,000 live births for the 5-year period before the survey | (Child mortality, 1990 JPFHS, 5) (Neonatal mortality, 1990 JPFHS, 21) (Infant mortality, 1990 JPFHS, 34) (Under-5 mortality, 1990 JPFHS, 39) (Child mortality, 1997 JPFHS, 6) (Neonatal mortality, 1997 JPFHS, 19) (Infant mortality, 1997 JPFHS, 29) (Under-5 mortality, 1997 JPFHS, 34) (Child mortality, 2002 JPFHS, 5) (Neonatal mortality, 2002 JPFHS, 16) (Infant mortality, 2002 JPFHS, 22) (Under-5 mortality, 2002 JPFHS, 27) (Child mortality, 2007 JPFHS, 2) (Neonatal mortality, 2007 JPFHS, 14) (Infant mortality, 2007 JPFHS, 19) (Under-5 mortality, 2007 JPFHS, 21) (Child mortality, 2009 JPFHS, 5) (Neonatal mortality, 2009 JPFHS, 15) (Infant mortality, 2009 JPFHS, 23) (Under-5 mortality, 2009 JPFHS, 28) (Child mortality, 2012 JPFHS, 4) (Neonatal mortality, 2012 JPFHS, 14) (Infant mortality, 2012 JPFHS, 17) (Under-5 mortality, 2012 JPFHS, 21) (Child mortality, 2017-18 JPFHS, 3) (Neonatal mortality, 2017-18 JPFHS, 11) (Infant mortality, 2017-18 JPFHS, 17) (Under-5 mortality, 2017-18 JPFHS, 19)\"\n }\n```", "flags": [ "" ], "info": "Provide a JSON formatted example of a typical instance in the dataset.", "scope": "periscope" }, { "title": "Data Splits", "level": 4, "content": "- `Train`: Training set, includes examples with 0 or more references.\n- `Validation`: Validation set, includes examples with 3 or more references.\n- `Test`: Test set, includes examples with 3 or more references.\n- `Ru`: Russian zero-shot set. Includes English and Russian examples (Russian is not includes in any of the other splits).", "flags": [ "" ], "info": "Describe and name the splits in the dataset if there are more than one.", "scope": "periscope" }, { "title": "Splitting Criteria", "level": 4, "content": "The same table across languages is always in the same split, i.e., if table X is in the test split in language A, it will also be in the test split in language B. In addition to filtering examples without transcribed table values, every example of the development and test splits has at least 3 references. \nFrom the examples that fulfilled these criteria, 100 tables were sampled for both development and test for a total of 800 examples each. A manual review process excluded a few tables in each set, resulting in a training set of 6,962 tables, a development set of 752 tables, and a test set of 763 tables.\n", "flags": [ "" ], "info": "Describe any criteria for splitting the data, if used. If there are differences between the splits (e.g., if the training annotations are machine-generated and the dev and test ones are created by humans, or if different numbers of annotators contributed to each example), describe them here.", "scope": "microscope" }, { "title": "", "level": 4, "content": "There are tables without references, without values, and others that are very large. The dataset is distributed as-is, but the paper describes multiple strategies to deal with data issues.", "flags": [ "" ], "info": "What does an outlier of the dataset in terms of length/perplexity/embedding look like?", "scope": "microscope" } ] } ] }, { "title": "Dataset in GEM", "level": 2, "subsections": [ { "title": "Rationale for Inclusion in GEM", "level": 3, "fields": [ { "title": "Why is the Dataset in GEM?", "level": 4, "content": "There is no other multilingual data-to-text dataset that is parallel over languages. Moreover, over 70% of references in the dataset require reasoning and it is thus of very high quality and challenging for models.", "flags": [ "" ], "info": "What does this dataset contribute toward better generation evaluation and why is it part of GEM?", "scope": "microscope" }, { "title": "Similar Datasets", "level": 4, "content": "yes", "flags": [ "" ], "info": "Do other datasets for the high level task exist?", "scope": "telescope" }, { "title": "Unique Language Coverage", "level": 4, "content": "yes", "flags": [ "" ], "info": "Does this dataset cover other languages than other datasets for the same task?", "scope": "periscope" }, { "title": "Difference from other GEM datasets", "level": 4, "content": "More languages, parallel across languages, grounded in infographics, not centered on Western entities or source documents", "flags": [ "" ], "info": "What else sets this dataset apart from other similar datasets in GEM?", "scope": "microscope" }, { "title": "Ability that the Dataset measures", "level": 4, "content": "reasoning, verbalization, content planning", "flags": [ "" ], "info": "What aspect of model ability can be measured with this dataset?", "scope": "periscope" } ] }, { "title": "GEM-Specific Curation", "level": 3, "fields": [ { "title": "Modificatied for GEM?", "level": 4, "content": "no", "flags": [ "" ], "info": "Has the GEM version of the dataset been modified in any way (data, processing, splits) from the original curated data?", "scope": "telescope" }, { "title": "Additional Splits?", "level": 4, "content": "no", "flags": [ "" ], "info": "Does GEM provide additional splits to the dataset?", "scope": "telescope" } ] }, { "title": "Getting Started with the Task", "level": 3, "fields": [ { "title": "Pointers to Resources", "level": 4, "content": "The background section of the [paper](https://arxiv.org/abs/2211.00142) provides a list of related datasets.", "flags": [ "" ], "info": "Getting started with in-depth research on the task. Add relevant pointers to resources that researchers can consult when they want to get started digging deeper into the task.", "scope": "microscope" }, { "title": "Technical Terms", "level": 4, "content": "- `data-to-text`: Term that refers to NLP tasks in which the input is structured information and the output is natural language.\n", "flags": [ "" ], "info": "Technical terms used in this card and the dataset and their definitions", "scope": "microscope" } ] } ] }, { "title": "Previous Results", "level": 2, "subsections": [ { "title": "Previous Results", "level": 3, "fields": [ { "title": "Metrics", "level": 4, "content": "`Other: Other Metrics`", "flags": [ "" ], "info": "What metrics are typically used for this task?", "scope": "periscope" }, { "title": "Other Metrics", "level": 4, "content": "`StATA`: A new metric associated with TaTA that is trained on human judgments and which has a much higher correlation with them.", "flags": [ "" ], "info": "Definitions of other metrics", "scope": "periscope" }, { "title": "Proposed Evaluation", "level": 4, "content": "The creators used a human evaluation that measured [attribution](https://arxiv.org/abs/2112.12870) and reasoning capabilities of various models. Based on these ratings, they trained a new metric and showed that existing metrics fail to measure attribution.", "flags": [ "" ], "info": "List and describe the purpose of the metrics and evaluation methodology (including human evaluation) that the dataset creators used when introducing this task.", "scope": "microscope" }, { "title": "Previous results available?", "level": 4, "content": "no", "flags": [ "" ], "info": "Are previous results available?", "scope": "telescope" } ] } ] }, { "title": "Dataset Curation", "level": 2, "subsections": [ { "title": "Original Curation", "level": 3, "fields": [ { "title": "Original Curation Rationale", "level": 4, "content": "The curation rationale is to create a multilingual data-to-text dataset that is high-quality and challenging.", "flags": [ "" ], "info": "Original curation rationale", "scope": "telescope" }, { "title": "Communicative Goal", "level": 4, "content": "The communicative goal is to describe a table in a single sentence.", "flags": [ "" ], "info": "What was the communicative goal?", "scope": "periscope" }, { "title": "Sourced from Different Sources", "level": 4, "content": "no", "flags": [ "" ], "info": "Is the dataset aggregated from different data sources?", "scope": "telescope" } ] }, { "title": "Language Data", "level": 3, "fields": [ { "title": "How was Language Data Obtained?", "level": 4, "content": "`Found`", "flags": [ "" ], "info": "How was the language data obtained?", "scope": "telescope" }, { "title": "Where was it found?", "level": 4, "content": "`Single website`", "flags": [ "" ], "info": "If found, where from?", "scope": "telescope" }, { "title": "Language Producers", "level": 4, "content": "The language was produced by USAID as part of the Demographic and Health Surveys program (https://dhsprogram.com/).", "flags": [ "" ], "info": "What further information do we have on the language producers?", "scope": "microscope" }, { "title": "Topics Covered", "level": 4, "content": "The topics are related to fertility, family planning, maternal and child health, gender, and nutrition.", "flags": [ "" ], "info": "Does the language in the dataset focus on specific topics? How would you describe them?", "scope": "periscope" }, { "title": "Data Validation", "level": 4, "content": "validated by crowdworker", "flags": [ "" ], "info": "Was the text validated by a different worker or a data curator?", "scope": "telescope" }, { "title": "Was Data Filtered?", "level": 4, "content": "not filtered", "flags": [ "" ], "info": "Were text instances selected or filtered?", "scope": "telescope" } ] }, { "title": "Structured Annotations", "level": 3, "fields": [ { "title": "Additional Annotations?", "level": 4, "content": "expert created", "flags": [ "quick" ], "info": "Does the dataset have additional annotations for each instance?", "scope": "telescope" }, { "title": "Number of Raters", "level": 4, "content": "11