[neat](https://github.com/qurator-spk/neat) is a simple, browser-based tool for editing and annotating text with named entities to produce a corpus for training/testing/evaluation. It can be used to add or correct named entity BIO-tags in a TSV file and to correct the token text or tokenization (e.g. due to OCR/segmentation errors).
[neat](https://github.com/qurator-spk/neat) is a simple, browser-based tool for editing and annotating text with named entities to produce a dataset for training/testing/evaluation. It can be used to add or correct named entity BIO-tags in a TSV file and to correct the token text or tokenization (e.g. due to OCR/segmentation errors).
[neat](https://github.com/qurator-spk/neat) is developed at the [Berlin State Library](https://staatsbibliothek-berlin.de/) for data annotation in the context of the [SoNAR-IDH](https://sonar.fh-potsdam.de/) project and the [QURATOR](https://qurator.ai/) project.
[neat](https://github.com/qurator-spk/neat) is developed at the [Berlin State Library](https://staatsbibliothek-berlin.de/) for data annotation in the [SoNAR-IDH](https://sonar.fh-potsdam.de/) project and the [QURATOR](https://qurator.ai/) project.
### 2. User Guide
### 2. User Guide
@ -40,7 +37,7 @@
[neat](https://github.com/qurator-spk/neat) runs locally as a pure HTML+JavaScript webpage in your web browser. No software needs to be installed, but JavaScript has to be enabled in the browser.
[neat](https://github.com/qurator-spk/neat) runs locally as a pure HTML+JavaScript webpage in your web browser. No software needs to be installed, but JavaScript has to be enabled in the browser.
#### 2.2. Installation
#### 2.2. Installation
Simply clone the repo using ``git clone https://github.com/qurator-spk/neat.git`` or download the [ZIP](https://github.com/qurator-spk/neat/archive/master.zip). Make sure you have at minimum ``neat.html`` and ``neat.js`` residing in a local directory, then it is sufficient to just open ``neat.html`` in a browser. Any fairly recent browser should work, but only Chrome and Firefox are tested.
Clone the repo using ``git clone https://github.com/qurator-spk/neat.git`` or download the [ZIP](https://github.com/qurator-spk/neat/archive/master.zip). Make sure you have ``neat.html`` and ``neat.js`` in the same directory and open ``neat.html`` in a browser. Any fairly recent browser should work, but only Chrome and Firefox are tested.
#### 2.3 Data format
#### 2.3 Data format
The data format is based on the format used in the [GermEval2014 Named Entity Recognition Shared Task](https://sites.google.com/site/germeval2014ner/data). Text is encoded as one token per line, with name spans encoded in the BIO-scheme, provided as tab-separated values:
The data format is based on the format used in the [GermEval2014 Named Entity Recognition Shared Task](https://sites.google.com/site/germeval2014ner/data). Text is encoded as one token per line, with name spans encoded in the BIO-scheme, provided as tab-separated values:
@ -76,9 +73,9 @@ No. TOKEN NE-TAG NE-EMB
```
```
For our purposes we extend this format by adding
For our purposes we extend this format by adding
* a fifth column for an ``ID`` for the outer ``NE-TAG`` from an authority file (in this case [Wikidata](https://www.wikidata.org/wiki/Wikidata:Main_Page) is used)
* a fifth column for an ``ID`` for the outer ``NE-TAG`` from an authority file
* column six for use as a variable ``url_id`` (see [Image Support](https://github.com/qurator-spk/neat/blob/master/README.md#28-image-support) for further details)
* column six for use as a variable ``url_id`` (see [Image Support](https://github.com/qurator-spk/neat/blob/master/README.md#27-image-support) for further details)
* finally, columns 7+ are used for storing ``left,right,top,bottom`` pixel coordinates for facsimile snippets
* finally, columns 7+ are used for storing ``left,right,top,bottom`` pixel coordinates for image snippets
The source data that is used for annotation are OCR results in [PAGE-XML](https://github.com/PRImA-Research-Lab/PAGE-XML) format. We provide a [Python tool](https://github.com/qurator-spk/page2tsv) that supports the transformation of [PAGE-XML](https://github.com/PRImA-Research-Lab/PAGE-XML) OCR files into the [TSV format](https://github.com/qurator-spk/neat/blob/master/README.md#23-data-format) required for use with [neat](https://github.com/qurator-spk/neat).
The source data that is used for annotation are OCR results in [PAGE-XML](https://github.com/PRImA-Research-Lab/PAGE-XML) format. We provide a [Python tool](https://github.com/qurator-spk/page2tsv) for the transformation of [PAGE-XML](https://github.com/PRImA-Research-Lab/PAGE-XML) OCR files into the [TSV format](https://github.com/qurator-spk/neat/blob/master/README.md#23-data-format) used by [neat](https://github.com/qurator-spk/neat).
#### 2.5 Provenance
#### 2.5 Keyboard-Navigation
The processing pipeline applied at the Berlin State Library comprises the follows steps:
OCR is based on [OCR-D](https://github.com/OCR-D)'s [ocrd_tesserocr](https://github.com/OCR-D/ocrd_tesserocr) which requires [Tesseract](https://github.com/tesseract-ocr/tesseract) **>= 4.1.0**. The [GT4HistOCR_2000000](https://ub-backup.bib.uni-mannheim.de/~stweil/ocrd-train/data/GT4HistOCR_2000000.traineddata) model, which is [trained](https://github.com/tesseract-ocr/tesstrain/wiki/GT4HistOCR) on the [GT4HistOCR](https://zenodo.org/record/1344132) corpus, is used. Further details are available in the [paper](https://arxiv.org/abs/1809.05501).
3. TSV Transformation
A simple [Python tool](https://github.com/qurator-spk/page2tsv) is used for the transformation of the OCR results in [PAGE-XML](https://github.com/PRImA-Research-Lab/PAGE-XML) to [TSV](https://github.com/qurator-spk/neat/blob/master/docs/README.md#23-data-format).
4. Tokenization
For tokenization, [SoMaJo](https://github.com/tsproisl/SoMaJo) is used.
5. Named Entity Recognition
For Named Entity Recognition, a [BERT-Base](https://github.com/google-research/bert) model was trained for noisy OCR texts with historical spelling variation. [sbb_ner](https://github.com/qurator-spk/sbb_ner) is using a combination of unsupervised training on a large (~2.3m pages) [corpus of German OCR](https://zenodo.org/record/3257041) in combination with supervised training on a small (47k tokens) [annotated corpus](https://github.com/EuropeanaNewspapers/ner-corpora/tree/master/enp_DE.sbb.bio). Further details are available in the [paper](https://corpora.linguistik.uni-erlangen.de/data/konvens/proceedings/papers/KONVENS2019_paper_4.pdf).
* left-click `<<` and `>>` to move 15 rows up or down
* left-click `<<` and `>>` to move 15 rows up or down
@ -179,14 +162,11 @@ For Named Entity Recognition, a [BERT-Base](https://github.com/google-research/b
* left-click the `POSITION` of a row and select `start-sentence` from the drop-down menu to start a new sentence
* left-click the `POSITION` of a row and select `start-sentence` from the drop-down menu to start a new sentence
#### 2.8 Image Support
#### 2.7 Image Support
Provided facsimile images are available online via the [iiif.io](https://iiif.io/) Image API, [neat](https://github.com/qurator-spk/neat) supports the embedding of facsimile snippets into its interface to help with data annotation and correction.
Provided facsimile images are available via the [iiif.io](https://iiif.io/) Image API, [neat](https://github.com/qurator-spk/neat) supports the embedding of image snippets into its interface to assist data annotation and correction. This requires that the PAGE-XML OCR contains word bounding boxes.
This further requires that OCR with word segmentation is applied to the image to determine bounding boxes for tokens.
The iiif-image-url contained in the source ``#`` can then be used as a replacement for ``url_id`` in combination with the token bounding boxes as ``left,right,top,bottom`` to obtain the facsimile snippet url and display the image in the leftmost column. Clicking on the facsimile snippet opens up a new tab with a larger context.
#### 2.9 Saving progress
#### 2.8 Saving progress
[neat](https://github.com/qurator-spk/neat) runs fully locally in the browser. Therefore it can not automatically save any changes you made to disk. You have to use the `Save Changes` button in order to so manually from time to time. If your browser automatically saves all downloads to your `Downloads` folder, you might want to configure it so that it instead prompts you where to save.
[neat](https://github.com/qurator-spk/neat) runs fully locally in the browser. Therefore it can not automatically save any changes you made to disk. You have to use the `Save Changes` button in order to so manually from time to time.
### 3. Annotation Guidelines
### 3. Annotation Guidelines
The most recent version of the [Annotation Guidelines](https://github.com/qurator-spk/neat/blob/master/Annotation_Guidelines.pdf) is included in this repository.
Clemens Neudecker
5 years ago
committed by
GitHub