An excerpt of the chapter on the infinitive mood.

The project

With Zeng Yi (曾怡) , I co-authored a textbook of ancient Greek in Mandarin. From three chapters of this textbook, I produced a natively accessible, enriched and interactive ePub3.3, as part of a dissertation supervised by Fabien Torre and Karolina Suchecka, at the Université de Lille.

The glossary at the end of the textbook

1. Converting into XML

The first step of the process consisted in:

  • Getting the flat XML from the styled .odt files
  • Modifying and refining the markup of the flat XML
    • XSLT
  • Adding a first level of linguistic markup with algorithms
    • XSLT
  • Extracting a glossary based on the word list of each chapter
    • XSLT
XML enriched with Perseus and Stanza

2. Enriching the files with NLP tools

The second step consisted in enriching the resulting XML files with Natural Language Processing tools [NLP].

  • With the Perseus module from the Stanza API, I added:
    • python3
    • an analysis in parts of speech (noun, verb, adjective…)
    • a morphological analysis (indicative present active…)
    • an analysis in universal dependencies (that is, a syntax analysis)
  • Connect each word to the glossary to show its meanings or its translations.
    • XSLT
Interface of the textbook in the epub format

3. Making the ePub and building a user interface

All that remained was to…

  • Make the born accessible ePub3.3
    • XSLT
  • Build a user interface to view the results of the NLP processing as optional hints
    • JavaScript
  • Embed this interface into the ePub