Home

Awesome

YEDDA Logo

YEDDA: A Lightweight Collaborative Text Span Annotation Tool

About:

YEDDA (the previous SUTDAnnotator) is developed for annotating chunk/entity/event on text (almost all languages including English, Chinese), symbol and even emoji. It supports shortcut annotation which is extremely efficient to annotate text by hand. The user only need to select text span and press shortcut key, the span will be annotated automatically. It also support command annotation model which annotates multiple entities in batch and support export annotated text into sequence text. Besides, intelligent recommendation and adminstrator analysis is also included in updated version. It is compatiable with all mainstream operating systems includings Windows, Linux and MacOS.

For more details, please refer to our paper (ACL2018:demo, best demo nomination).

This GUI annotation tool is developed with tkinter package in Python.

System required: Python 3.6 or above

Author: Jie Yang, Assistant Professor at Zhejiang University, China.

Interface:

It provides both annotator interface for efficient annotatation and admin interface for result analysis.

Use as an annotator ?

This tool supports two ways of annotation (annotated text format [@the text span＃Location*]):

Intelligent recommendation:

The annotated results will be stored synchronously. Annotated file is located at the same directory with origin file with the name of "origin name + .ann" Please also note that the shortcut map can be switched seamlessly in the right bottom drop-down list

Use as an administrator ?

YEDDA provides a simple interface for administartor to evaluate and analyze annotation quality among multiple annotators. After collected multiple annotated *.ann files from multiple annotators (annotated on same plain text), YEDDA can give two toolkits to monitor the annotation quality: multi-annotator analysis and pairwise annotators comparison.

alt text

alt text

Important features:

  1. Type ctrl + z will undo the most recent modification
  2. Put cursor within an entity span, press shortcut key (e.g. x) to update label (binded with x) of the entity where cursor is belonging. (q for remove the label)
  3. Selected the annotated text, such as [@美国＃Location*], then press q, the annotated text will be recoverd to unannotate format (i.e. "美国").
  4. Change label directly, select entity content or put cursor inside the entity span (such as [@美国＃Location*]), then press x, the annotated text will change to new label mapped with shortcut x (e.g. [@美国#Organization*]).
  5. Confirm or remove recommended entity: put cursor inside of the entity span and press y (yes) or q (quit).
  6. In the command entry, just type Enter without any command, the cursor in text will move to the head of next line. (You can monitor this through "Cursor").
  7. The "Cursor" shows the current cursor position in text widget, with row and col represent the row and column number, respectively.
  8. Export button will export the ".ann" file as a identity name with ".anns" in the same directory. The exported file list the content in sequence format. In the source code, there is a flag self.seged which controls the exported bahaviour. a). If your sentences are consist of words seperated with space (e.g. segmentated Chinese and English), then you may set self.seged=True. b). If your sentences are consist of characters without space (e.g. unsegmentated Chinese text), set self.seged=False. Another flag self.tagScheme controls the exporting format, the exported ".anns" will use the BMES format if this flag is set to "BMES", otherwise the exported file is formatted as "BIO".

Cite:

If you use YEDDA for research, please cite our ACL paper as follows:

@article{yang2017yedda,  
 title={YEDDA: A Lightweight Collaborative Text Span Annotation Tool},  
 author={Yang, Jie and Zhang, Yue and Li, Linwei and Li, Xingxuan},  
 booktitle={Proceedings of the 56th Annual Meeting of the Association for Computational Linguistics},
 url={http://aclweb.org/anthology/P18-4006},
 year={2018}  
}

Updating...