Other Open Access

OCR models for Occitan (standard spelling)

Marianne Vergez-Couret


MARC21 XML Export

<?xml version='1.0' encoding='UTF-8'?>
<record xmlns="http://www.loc.gov/MARC21/slim">
  <leader>00000nam##2200000uu#4500</leader>
  <datafield tag="653" ind1=" " ind2=" ">
    <subfield code="a">OCR module, Tesseract, Jochre, Occitan, OCR</subfield>
  </datafield>
  <controlfield tag="005">20190108104917.0</controlfield>
  <controlfield tag="001">2533877</controlfield>
  <datafield tag="856" ind1="4" ind2=" ">
    <subfield code="s">5040156</subfield>
    <subfield code="z">md5:670f6b06177a947140dac65acb1ad123</subfield>
    <subfield code="u">https://zenodo.org/record/2533877/files/JOCHRE_2015.zip</subfield>
  </datafield>
  <datafield tag="856" ind1="4" ind2=" ">
    <subfield code="s">3182964</subfield>
    <subfield code="z">md5:2496c68c163afe1a8e473d81ce0d802d</subfield>
    <subfield code="u">https://zenodo.org/record/2533877/files/TESS_2015.traineddata</subfield>
  </datafield>
  <datafield tag="542" ind1=" " ind2=" ">
    <subfield code="l">open</subfield>
  </datafield>
  <datafield tag="260" ind1=" " ind2=" ">
    <subfield code="c">2019-01-08</subfield>
  </datafield>
  <datafield tag="909" ind1="C" ind2="O">
    <subfield code="p">user-restaure</subfield>
    <subfield code="o">oai:zenodo.org:2533877</subfield>
  </datafield>
  <datafield tag="100" ind1=" " ind2=" ">
    <subfield code="u">Université de Poitiers</subfield>
    <subfield code="a">Marianne Vergez-Couret</subfield>
  </datafield>
  <datafield tag="245" ind1=" " ind2=" ">
    <subfield code="a">OCR models for Occitan (standard spelling)</subfield>
  </datafield>
  <datafield tag="980" ind1=" " ind2=" ">
    <subfield code="a">user-restaure</subfield>
  </datafield>
  <datafield tag="540" ind1=" " ind2=" ">
    <subfield code="u">https://creativecommons.org/licenses/by/4.0/legalcode</subfield>
    <subfield code="a">Creative Commons Attribution 4.0 International</subfield>
  </datafield>
  <datafield tag="650" ind1="1" ind2="7">
    <subfield code="a">cc-by</subfield>
    <subfield code="2">opendefinition.org</subfield>
  </datafield>
  <datafield tag="520" ind1=" " ind2=" ">
    <subfield code="a">&lt;p&gt;This dataset provides trained Tesseract (&lt;a href="https://github.com/tesseract-ocr/tesseract"&gt;https://github.com/tesseract-ocr/tesseract&lt;/a&gt;) and Jochre (&lt;a href="https://github.com/urieli/jochre"&gt;https://github.com/urieli/jochre&lt;/a&gt;) OCR models for Occitan ( for the standard spelling and two dialects, Gascon and Lengadocian). These models were developed in the context of the RESTAURE project, funded by the French ANR.&amp;nbsp;&lt;/p&gt;

&lt;p&gt;Two models are provided. They were presented in the following article &lt;a href="http://hal.archives-ouvertes.fr/hal-01252241"&gt;https://hal.archives-ouvertes.fr/hal-01252241&lt;/a&gt; and also re-evaluated for the creation of another corpus in &lt;a href="https://www.openscience.fr/Constitution-et-annotation-d-un-corpus-ecrit-de-contes-et-recits-en-occitan"&gt;https://www.openscience.fr/Constitution-et-annotation-d-un-corpus-ecrit-de-contes-et-recits-en-occitan&lt;/a&gt;.&lt;/p&gt;

&lt;p&gt;The first model for Jochre, JOCHRE_2015, has been trained for Jochre 1.1.2b. The training images and corresponding texts were manually annotated using a Jochre online platform (excerpts from 7 different printed works, totalling about 20,000 words)&lt;/p&gt;

&lt;p&gt;The second model for Tesseract, TESS_2015, was trained using the jTessBoxEditor tool (&lt;a href="http://vietocr.sourceforge.net/training.html"&gt;http://vietocr.sourceforge.net/training.html&lt;/a&gt;), Version 1.4 (2 May 2015), based on images automatically generated from the training texts (the one used for Jochre). The generation of the images used a 36pt font size, and two fonts were used (Arial and Times New Roman), with their normal and italic variants. The Tesseract model can be used with Tesseract 3.0x.&lt;/p&gt;

&lt;p&gt;List of words was also used for those two trainings. We conflated Occitan words found in several lexicons, dictionaries and corpora for the two dialects, Gascon and Lengadocian:&lt;/p&gt;

&lt;ul&gt;
	&lt;li&gt;Lexicon extracted from 60 literary works (from 29 different authors) gathered in the BaTel&amp;Ograve;c project.&lt;/li&gt;
	&lt;li&gt;Dictonary entries from &lt;em&gt;Dictionnaire Fran&amp;ccedil;ais/Occitan Gascon Toulousain&lt;/em&gt; de Nicolau Rei B&amp;egrave;thv&amp;eacute;der, 2004, IEO Edicions&lt;/li&gt;
	&lt;li&gt;Dictonary entries from &lt;em&gt;Dictionnaire Fran&amp;ccedil;ais/Occitan&lt;/em&gt; de Cristian Laus, 2004, IEO/IDECO&lt;/li&gt;
	&lt;li&gt;Dictonary entries from &lt;em&gt;Dictionnaire Fran&amp;ccedil;ais/Occitan (Gascon)&lt;/em&gt; de Miqu&amp;egrave;u Grosclaude, Gilab&amp;egrave;rt Nari&amp;ograve;o e Patric Guilhemjoan, 2007, Per Noste Edicions&lt;/li&gt;
	&lt;li&gt;Conjugated forms from Verb&amp;rsquo;&amp;Ograve;c (designed by the &lt;em&gt;Congr&amp;egrave;s permanent de la lenga occitana&lt;/em&gt; (&lt;a href="http://www.locongres.org"&gt;http://www.locongres.org&lt;/a&gt;))&lt;/li&gt;
	&lt;li&gt;List of proper nouns extracted from the Apertium (free/open-source machine translation platform) Occitan lexicon.&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;The jochre model can be used with the Jochre software (&lt;a href="https://github.com/urieli/jochre"&gt;https://github.com/urieli/jochre&lt;/a&gt;). See also Jochre wiki (https://github.com/urieli/jochre/wiki).&lt;/p&gt;

&lt;p&gt;The Tesseract models can be used&amp;nbsp; for instance using the gImageReader tool (&lt;a href="https://github.com/manisandro/gImageReader"&gt;https://github.com/manisandro/gImageReader&lt;/a&gt;), which provides a graphical user interface for the Tesseract tool.&amp;nbsp;&lt;/p&gt;

&lt;p&gt;When evaluated against the same test corpus (four extracts from four different authors from two dialects, Gascon and Lengadocian), the Jochre model achieves better performance levels.&lt;/p&gt;

&lt;p&gt;&amp;nbsp;&lt;/p&gt;</subfield>
  </datafield>
  <datafield tag="773" ind1=" " ind2=" ">
    <subfield code="n">doi</subfield>
    <subfield code="i">isVersionOf</subfield>
    <subfield code="a">10.5281/zenodo.2533876</subfield>
  </datafield>
  <datafield tag="024" ind1=" " ind2=" ">
    <subfield code="a">10.5281/zenodo.2533877</subfield>
    <subfield code="2">doi</subfield>
  </datafield>
  <datafield tag="980" ind1=" " ind2=" ">
    <subfield code="a">other</subfield>
  </datafield>
</record>
145
22
views
downloads
All versions This version
Views 145145
Downloads 2221
Data volume 90.5 MB85.4 MB
Unique views 136136
Unique downloads 2019

Share

Cite as