Minutes |
LEXIDMA meeting 19 October 2020, 14:30-16:00 CEST
1. Administrative
- roll call - Tomaz, Simon, Iztok, John, David, Milos, Michal, Ilan
- appoint scribe - Ilan
- approve minutes from previous meeting 21st September
https://markmail.org/message/3o7mvduysamjobqz
- approved.
- Add to AOB – review GANTT
2. Material
- Michal prepared report on subsensing in dictionaries
https://markmail.org/thread/fz46xz362r5n2fer [mailing list discussion]
- was uploaded by David to the repository for future reference
https://markmail.org/thread/fz46xz362r5n2fer [mailing list discussion]
- serving as a proposal for remodeling
- has the benefit of flattening senses rather than hierarchical sub-senses
- interoperable by machines
- would not be losing any information available in the dictionaries studied
- John: DWDS example is odd – Michal agrees there should not be separate entries
- David: is relation-based subsensing recursion, to avoid having an element contain itself
- Michal: avoid illogical recursion by allowing subsensing
- David: could have two different entries, one having
- Ilan: when ready, expand with regard to nested entries and MWEs (as subsenses)
- Three main patterns: senses inside senses, the child sense is MWE, entry inside entry (incl homographs)
- take care of all of them by flattening
- David: should decide who is primary user – machine or human
- Miloš: homographs need not be indicated explicitly – no clear distinction of homonymy and polysemy – lexicographer may decide either way, the machine can deal with either
- Simon: possible to establish relation between the two
- Miloš: Can infer homographs from similar spelling
- Simon: changing the world of lexicography from book-based logic, so keep the possibility of labeling relations – can label two homonym entries as homonymic
- Michal: if we follow the same strategy, we become truly members of the digital era
- Michal will write similar proposal including homographs/homonyms, nesting, MWEs
- David: we need to be language-agnostic in this sense
- MWE – making it relational
- Simon: underdefined, not typically headwords, should not be modelled as subsenses
- David: don’t agree, depends whether has compositional sense or not
- Simon: agglutinative language has other problem, as could have millions of headwords; we are talking about move from hierarchical book model – whatever one considers as headword, or as something that should be explained with some kind of lexicographic identity, it can be a headword; otherwise it is used as part of the sense
- David: we need a more comprehensive proposal, include examples and describe mechanics
- Michal confirms accepting to write wider proposal
3. Closing Admin AOB
- Gantt chart – we were on track with administrative things, launched the TC by end of 2019
- Had pre-assumption that working draft of deliverable starts right away, but we needed design discussions to precede
- We are 8 months late with regard to the Gantt chart
- David: Will need a year for drafting the standard
- Simon suggests deadline to start public reviews before summer 2021
- Miloš agrees with Ilan that the discussion of the underlying principals was substantial and needed fairly long time but now we can speed up the process
- Heads up for spec formatting decision next time
- David brings from backburner having to determine the formatting question – Oasis style or ISO style
- David send the two variants to the mailing list
- David will present option to discuss next time, summarize and put on agenda next time
- Political thing – fasttrack in Oasis or longterming with ISO
- Simon and David are editors, everyone can contribute
Adjourned 1500 WEST |