Datasets:
annotations_creators:
- machine-generated
language_creators:
- found
language:
- en
license:
- cc-by-4.0
multilinguality:
- monolingual
size_categories:
- 1M<n<10M
source_datasets:
- original
task_categories:
- token-classification
task_ids:
- part-of-speech
paperswithcode_id: twitter-pos-vcb
pretty_name: Twitter PoS VCB
Dataset Card for "twitter-pos-vcb"
Table of Contents
- Dataset Description
- Dataset Structure
- Dataset Creation
- Considerations for Using the Data
- Additional Information
Dataset Description
- Homepage: https://gate.ac.uk/wiki/twitter-postagger.html
- Repository: https://github.com/GateNLP/gateplugin-Twitter
- Paper: https://aclanthology.org/R13-1026.pdf
- Point of Contact: Leon Derczynski
- Size of downloaded dataset files: 4.51 MiB
- Size of the generated dataset: 26.88 MB
- Total amount of disk used: 31.39 MB
Dataset Summary
Part-of-speech information is basic NLP task. However, Twitter text is difficult to part-of-speech tag: it is noisy, with linguistic errors and idiosyncratic style. This data is the vote-constrained bootstrapped data generate to support state-of-the-art results.
The data is about 1.5 million English tweets annotated for part-of-speech using Ritter's extension of the PTB tagset. The tweets are from 2012 and 2013, tokenized using the GATE tokenizer and tagged jointly using the CMU ARK tagger and Ritter's T-POS tagger. Only when both these taggers' outputs are completely compatible over a whole tweet, is that tweet added to the dataset.
This data is recommend for use a training data only, and not evaluation data.
For more details see https://gate.ac.uk/wiki/twitter-postagger.html and https://aclanthology.org/R13-1026.pdf
Supported Tasks and Leaderboards
Languages
English, non-region-specific. bcp47:en
Dataset Structure
Data Instances
An example of 'train' looks as follows.
Data Fields
The data fields are the same among all splits.
twitter_pos_vcb
id: astringfeature.tokens: alistofstringfeatures.pos_tags: alistof classification labels (int). Full tagset with indices:
Data Splits
| name | tokens | sentences |
|---|---|---|
| twitter-pos-vcb | 1 543 126 | 159 492 |
Dataset Creation
Curation Rationale
Source Data
Initial Data Collection and Normalization
Who are the source language producers?
Annotations
Annotation process
Who are the annotators?
Personal and Sensitive Information
Considerations for Using the Data
Social Impact of Dataset
Discussion of Biases
Other Known Limitations
Additional Information
Dataset Curators
Licensing Information
Creative Commons Attribution 4.0 (CC-BY)
Citation Information
@inproceedings{derczynski2013twitter,
title={Twitter part-of-speech tagging for all: Overcoming sparse and noisy data},
author={Derczynski, Leon and Ritter, Alan and Clark, Sam and Bontcheva, Kalina},
booktitle={Proceedings of the international conference recent advances in natural language processing ranlp 2013},
pages={198--206},
year={2013}
}
Contributions
Author uploaded (@leondz)