Word-Lemma-Lists and Lexica used by PTA

The folder sources contains the source files of the used lexica; for more information see the README.md in this folder.

The folder scripts contains the Python scripts used to generate all files.

Greek

The repository has the following files at the moment:

• Bailly2020.3a.json, LSJ.json, Pape.json, TBESG.json Greek dictionaries compiled from respective source files in sources.
• pta_lexicon_grc.json:
  • compiled from LSJ, TBESG, and Pape
  • has: lemma – grc_eng – grc_eng2 – grc_deu
  • grc_eng = LSJ, grc_eng2 = TBESG, grc_deu = Pape
  • grc_eng and grc_deu are lists, as there are homonymous lemmata.
  • If there is no entry in one of the dictionaries, the entry is empty.
• The folder pta_lexicon_grc contains xml-version of the above
• wordlemma_grc_cltk.json:
  • result of lemmatizing all Greek texts in in pta_data; it currently has 133.438 entries. Lemmatization was done using the Classical Language Toolkit (CLTK).
  • has word - lemma - POS - morphology (according to Universal Dependencies (UD) project)
• wordlemma_grc.json (outdated):
  • result of lemmatizing part of the texts in in pta_data; it has 42.346 entries. Lemmatization was done using the Morpheus morphological analysis engine used at morph.perseids.org.
  • has word – lemma – morphology
  • words which have not been lemmatized (for whatever reason), are not in the file.
• wordlemma_grc.xml (outdated):
  • xml-version of the file above
• wordlemma_grc_diogenes.json:
  • morphology data from Diogenes; Greek is converted to utf-8 (from Betacode).
  • has word - lemma (list of possible morphology)
• JSON-versions of the lexica in the source-folder, adapted for use in PTA, the folder pta_dictionaries contains xml-versions of these.

Latin

• Georges.json: compiled from respective source file in sources.

• LewisShort.json: tbd

• TLL.json:

  • built from https://publikationen.badw.de/de/api/thesaurus/html-xml/thesaurus/index.json"
  • has lemma - url of entry in THESAVRVS LINGVAE LATINAE Open Access

• wordlemma_lat_cltk.json:

  • result of lemmatizing all Latin texts in in pta_data; it currently has 3022 entries. Lemmatization was done using the Classical Language Toolkit (CLTK).
  • has word - lemma - POS - morphology (according to Universal Dependencies (UD) project)

• wordlemma_lat_diogenes.json:

  • morphology data from Diogenes
  • has word - lemma (list of possible morphology)