Laboratoire ICAR » Plateforme TXM

Texts are possibly composed of several textual planes:

• the basic text, or primary text or main body of text
• section titles (it can be a choice to consider those as not participating to the basic text content - for example when the title is added by the scientific editor)
• citations, possibly in different languages (it can be a choice to differenciate those from the basic text - for example to build the lexicon of the basic text)
• speaker discourses (it can be a choice to consider or to ignore the discourse of specific speakers)
• footnotes text, possibly in different languages and possibly embeded (hierarchy of critical apparatus)

We must help the user to use various textual planes for searching or counting.

Solution:

• make import modules aware of textual planes and design corresponding import parameters
  • drive NLP tools with respect to languages found in textual planes
• design UI to select which textual planes are used
• relate textual planes to sub-corpus management (for example through the 'sequence of textual planes' CQP encoding strategy)
• relate textual planes to 'Text edition with source text not in corpus' management, see #968