Home › Datasets › modality › Texts
Texts datasets
archive 2025-07-283,130 datasets carry the modality tag "Texts", ordered by the archive's paper count. Page 9 of 66: 48 shown of 3,130. Facet routes are this site's own (the archive records the tag string, not a page).
The archive holds 12,214 dataset rows; 12,172 are listed. 6 are withheld from every listing and count here as vandalised before snapshot (6 with contact-centre spam in the title, 0 with a spam description on a row that has no homepage, no paper and no papers counted; none with more than 1 paper, 0 with a benchmark), listed in withheld.json; 1 listed row carries a vandalised description, withheld on its page. This gate never withholds a row with a homepage or a paper that resolves, and a clean description; the content rules below withhold a row whose name is spam whatever else it carries. The gate is a phrase list: these are the rows it caught, not a claim that the rest is clean. Before that gate, the site's content rules withhold 36 more rows (invite-code, gambling, travel-booking, contact-centre and similar spam in the name or on a row with nothing real behind it); they have no page and are listed in withheld.json.
Filter 39 modality tags shown of 39, by dataset count; the full filter by modality, task and language is on /datasets
All 12,172
Images 3,239Texts 3,130Videos 1,014Audio 480Medical 3943D 380Time series 294Graphs 282Tabular 267Speech 197RGB-D 190Environment 147Point cloud 134Biomedical 122LiDAR 94RGB Video 87Tracking 77Biology 70Actions 673d meshes 64Tables 50Music 47EEG 44Hyperspectral images 44Stereo 44MRI 39Physics 32Interactive 29Dialog 25Midi 226D 17Replay data 11Financial 10Ranking 10Cad 9fMRI 7Parallel 6Lyrics 2PSG 2
Texts datasets 385–432 of 3,130
GeoQA (Geometric Question Answering)
GeoQA is a dataset for automatic geometric problem solving containing 5,010 geometric problems with corresponding annotated programs, which illustrate the solving process of the given problems Compared with another publicly available…
46 papers · 1 benchmark
Legal General Language Understanding Evaluation (LexGLUE) benchmark is a collection of datasets for evaluating model performance across a diverse set of legal NLU tasks in a standardized way.
46 papers · 1 benchmark
MAMS (Multi Aspect Multi-Sentiment)
MAMS is a challenge dataset for aspect-based sentiment analysis (ABSA), in which each sentences contain at least two aspects with different sentiment polarities.
46 papers · 1 benchmark
UDC (Ubuntu Dialogue Corpus)
Ubuntu Dialogue Corpus (UDC) is a dataset containing almost 1 million multi-turn dialogues, with a total of over 7 million utterances and 100 million words.
46 papers · 8 benchmarks
CoSQL (Conversational Text-to-SQL Challenge)
CoSQL is a corpus for building cross-domain, general-purpose database (DB) querying dialogue systems.
45 papers · 1 benchmark
DART is a large dataset for open-domain structured data record to text generation.
45 papers · 3 benchmarks
GEM (Generation, Evaluation, and Metrics)
Generation, Evaluation, and Metrics (GEM) is a benchmark environment for Natural Language Generation with a focus on its Evaluation, both through human annotations and automated Metrics.
45 papers · 1 benchmark
A new large-scale geometry problem-solving dataset - 3,002 multi-choice geometry problems - dense annotations in formal language for the diagrams and text - 27,213 annotated diagram logic forms (literals) - 6,293 annotated text logic forms…
45 papers · 1 benchmark
A new dataset of 1,001 human-human dialogs for movie recommendation with measures for successful recommendations.
45 papers · 0 benchmarks
MLSUM (MultiLingual SUMmarization)
A large-scale MultiLingual SUMmarization dataset.
45 papers · 4 benchmarks
RSTPReid (Real Scenario Text-based Person Re-identification)
RSTPReid contains 20505 images of 4,101 persons from 15 cameras.
45 papers · 2 benchmarks
EmotionLines contains a total of 29245 labeled utterances from 2000 dialogues.
44 papers · 1 benchmark
How2Sign (A Large-scale Multimodal Dataset for Continuous American Sign Language)
The How2Sign is a multimodal and multiview continuous American Sign Language (ASL) dataset consisting of a parallel corpus of more than 80 hours of sign language videos and a set of corresponding modalities including speech, English…
44 papers · 3 benchmarks
OCNLI (Original Chinese Natural Language Inference)
OCNLI stands for Original Chinese Natural Language Inference.
44 papers · 0 benchmarks
The SCUT-CTW1500 dataset contains 1,500 images: 1,000 for training and 500 for testing.
44 papers · 3 benchmarks
TurkCorpus, a dataset with 2,359 original sentences from English Wikipedia, each with 8 manual reference simplifications.
44 papers · 1 benchmark
For understanding multimodal language used in expressing humor.
44 papers · 0 benchmarks
Dataset contains 33,010 molecule-description pairs split into 80\%/10\%/10\% train/val/test splits.
43 papers · 4 benchmarks
QA-SRL was proposed as an open schema for semantic roles, in which the relation between an argument and a predicate is expressed as a natural-language question containing the predicate (“Where was someone educated?”) whose answer is the…
43 papers · 0 benchmarks
ShARC (Shaping Answers with Rules through Conversation)
ShARC is a Conversational Question Answering dataset focussing on question answering from texts containing rules.
43 papers · 0 benchmarks
WebQA, is a new benchmark for multimodal multihop reasoning in which systems are presented with the same style of data as humans when searching the web: Snippets and Images.
43 papers · 0 benchmarks
gRefCOCO is the first large-scale Generalized Referring Expression Segmentation dataset that contains multi-target, no-target, and single-target expressions.
43 papers · 2 benchmarks
BUCC (Building and Using Comparable Corpora)
The BUCC mining task is a shared task on parallel sentence extraction from two monolingual corpora with a subset of them assumed to be parallel, and that has been available since 2016.
42 papers · 4 benchmarks
DuoRC contains 186,089 unique question-answer pairs created from a collection of 7680 pairs of movie plots where each pair in the collection reflects two versions of the same movie.
42 papers · 1 benchmark
We release E-commerce Dialogue Corpus, comprising a training data set, a development set and a test set for retrieval based chatbot.
42 papers · 1 benchmark
EmoContext consists of three-turn English Tweets.
42 papers · 1 benchmark
English Web Treebank is a dataset containing 254,830 word-level tokens and 16,624 sentence-level tokens of webtext in 1174 files annotated for sentence- and word-level tokenization, part-of-speech, and syntactic structure.
42 papers · 0 benchmarks
IconQA (Icon Question Answering)
Current visual question answering (VQA) tasks mainly consider answering human-annotated questions for natural images in the daily-life context.
42 papers · 1 benchmark
PearRead is a dataset of scientific peer reviews.
42 papers · 0 benchmarks
SCROLLS (Standardized CompaRison Over Long Language Sequences)
SCROLLS (Standardized CompaRison Over Long Language Sequences) is an NLP benchmark consisting of a suite of tasks that require reasoning over long texts.
42 papers · 1 benchmark
Amazon Product Data (rithik)
This dataset contains product reviews and metadata from Amazon, including 142.8 million reviews spanning May 1996 - July 2014.
41 papers · 5 benchmarks
A creative writing task where the input is 4 random sentences and the output should be a coherent passage with 4 paragraphs that end in the 4 input sentences respectively.
41 papers · 0 benchmarks
QVHighlights (Query-based Video Highlights)
The Query-based Video Highlights (QVHighlights) dataset is a dataset for detecting customized moments and highlights from videos given natural language (NL).
41 papers · 4 benchmarks
A ChatGPT-Assisted Weakly-Labelled Audio Captioning Dataset for Audio-Language Multimodal Research.
41 papers · 0 benchmarks
BLURB (Biomedical Language Understanding and Reasoning Benchmark)
BLURB is a collection of resources for biomedical natural language processing.
40 papers · 2 benchmarks
Contains around 200K dialogs with a total of 1.6M turns.
40 papers · 0 benchmarks
40 papers · 0 benchmarks
Samanantar is the largest publicly available parallel corpora collection for Indic languages: Assamese, Bengali, Gujarati, Hindi, Kannada, Malayalam, Marathi, Oriya, Punjabi, Tamil, Telugu.
40 papers · 0 benchmarks
SciCite is a dataset of citation intents that addresses multiple scientific domains and is more than five times larger than ACL-ARC.
40 papers · 3 benchmarks
We propose the first question-answering dataset driven by STEM theorems.
40 papers · 1 benchmark
ALCE (Automatic LLMs' Citation Evaluation)
ALCE is a benchmark for Automatic LLMs' Citation Evaluation.
39 papers · 0 benchmarks
Break is a question understanding dataset, aimed at training models to reason over complex questions.
39 papers · 0 benchmarks
BookSum is a collection of datasets for long-form narrative summarization.
39 papers · 2 benchmarks
The Memetracker corpus contains articles from mainstream media and blogs from August 1 to October 31, 2008 with about 1 million documents per day.
39 papers · 1 benchmark
WikiMovies is a dataset for question answering for movies content.
39 papers · 0 benchmarks
AISHELL-3 is a large-scale and high-fidelity multi-speaker Mandarin speech corpus which could be used to train multi-speaker Text-to-Speech (TTS) systems.
38 papers · 0 benchmarks
BIOSSES (Biomedical Semantic Similarity Estimation System)
The BIOSSES data set comprises total 100 sentence pairs all of which were selected from the "TAC2 Biomedical Summarization Track Training Data Set" .
38 papers · 2 benchmarks
Paper counts and descriptions are the archive's, frozen 2025-07-28; no citation counts, no stars, no trending. Sorting by "most cited" or "newest" was a live-site feature the archive does not carry.