Meeting_2008-04-28

Meeting setup

  • Date: 28.4.2008
  • Time: 12.30 Norw. time
  • Place: Internet
  • Tools: SubEthaEdit, iChat/Skype

Agenda

Cf. one of the following, depending on context:

  • the upper bar of the SEE window (provided you use the JSPWiki syntax mode)
  • the TOC in Forrest-rendered output, like HTML and PDF

Opening, agenda review, participants

Opened at 10: 06.

Present: Børre, Jovsset, Per-Eric, Sjur, Thomas, Trond

Absent: Tomi

Agenda accepted as is.

Updated task status since last meeting

Børre

  • Hunspell lexicon conversion
    • found out how to convert the four big lexicons to hunspell. No compounding info yet, though.
  • prepare migration to svn (with Sjur, Trond)
    • have converted our cvs to svn, svn+ssh access works for me. Have had problems setting up apache authentication
  • release hunspell public beta at the end of April (with Sjur)
  • review hunspell lexicon branch (with Thomas), merge with trunk if ok
    • Thomas is working on it
  • try to repair G5 accounts for iCal Server
    • not done
  • fix bugs!

Lene

  • Ped project

Maaren

  • Put the list of possible sma corpus sources into a document

Per-Eric

  • try to find other authors who have smj texts digitaly, send them contracts
    • Nothing done
  • Work with missing list from texts written by Sigga Tuolja Sandström.
    • Worked and still working
  • Work with missing list same_dutkama_pgr.txt
    • Nothing done yet
  • Work with missing list sameriekta_tjoahkkagæsos.txt
    • Nothing done yet
  • Keep the contact with Ulf-Stefan Winka who has many more smj texts to add.
    • Nothing done
  • fix bugs!
    • Nothing done

Saara

  • add new XSL/XML headers for proofing test docs
  • Set up ways of adding meta-information for proofing correct corpus docs (source info, used in testing or not, added to lexicon or not)
  • discuss more parallel texts

Sjur

  • gather sma texts
    • several new contacts during the trip last week, two half-finished contracts on its way
  • name db/risten.no
    • nothing - was travelling last week
  • make an improved sma project plan
    • nothing - was travelling last week
  • publish corpus contracts and project infra as open-source on NoDaLi-sta
    • nothing - was travelling last week
  • prepare migration to svn (with Børre, Trond)
    • nothing - was travelling last week
  • release hunspell public beta at the end of April (with Børre)
    • nothing - was travelling last week
  • update the Changes document
    • nothing - was travelling last week
  • follow-up on some Polderland-related bugs: 621, 630, 652, 656
    • nothing - was travelling last week
  • test latest hunspell lexicon
    • forgot it: (
  • InDesign documentation
    • nothing - was travelling last week
  • fix bugs!
  • other things:
    • made a first version of a MacOS X dictionary for sme-smj and sme-nob based on the results of the UiT projects. Stll missing: morphology

Thomas

  • look at test cases still not behaving properly
    • nothing this week
  • review hunspell lexicon branch with Børre
    • working on this
  • fix bugs!
    • nothing this week

Tomi

  • Hunspell lexicon conversion
    • helped Børre with this
  • document how compounding is controlled in the PLX conversion
    • not done
  • fix double hyphen bugs
    • not done
  • Make a pedagogical speller (after MA thesis is delivered)
    • not done
  • fix bugs!

Trond

  • Help Jovsset with vislcg3 and sma
  • Set up Jabber for Lene, Kimme, Saara
  • Prepare svn migration (with Sjur, Børre)
  • fix bugs!.

Pedagogical software online

Working on it: Saara has a demo UI ready for vocabulary games, but not online yet. Lene has worked a lot on the word lists.

TODO:

  • get the content ready (Lene)
  • implement the UI and functionality (Saara)
  • get an easy-to-remember URL (UiT/IT)
  • More thorough skin, layout, ... (External person within the Ped team, Internal forrest expert) This we will postponed until later
  • Make a pedagogical speller (Tomi when finished with his MA thesis)
    • Turn off peripheral compounds (numbers, acros, perhaps names)
    • Increase editing distance by one for suggestions? Only possible with limited compounding

Documentation

TODO:

  • start to reorganise the documentation (Børre, Sjur, Trond)

Corpus gathering

Trond, Sjur, Jovsset were travelling last week, and got several contacts for sma corpus content. To be followed up.

TODO:

  • follow up on sma corpus texts (Sjur, Jovsset)
  • follow-up on the smj texts from Kurt Tore ( Per-Eric)
  • get texts from Sigga Tuolja Sandstrøm, possibly through Ulf-Stefan Winka (contract is ok now) (Per-Eric)
  • other contacts: Nord-Salten avis (Børge Strandskog), Lena Davidsson daughter to Lars-Matto Tuolja
  • Put the list of possible corpus sources into a document gt/doc/lang/sma/sma-corpus-plan.jspwiki ( Maaren)
  • give contract with blank fields to Per-Eric ( Børre)

Future plans, directions and ideas

See a separate document in plan/strat/5year.jspwiki.

Infrastructure

Changes in the lexicon cause big problems for Lene - she is using up to half her time to "fix" her rules to adapt to these changes. Types of changes:

  • tag changes
  • word changes (from lexicalised to compound)
  • classification changes (from abbr to acro, or similar)

Conclusions:

  • public pages and services should only be updated manually, and when we know the update is working
  • the whole product line must be more robust, such that small, local changes do not cause widespread errors
  • we need a larger test battery to run regularly (every night, before larger changes
  • a set of smaller tests that can be run automatically as part of the commit process

To accomodate future enhancements in different directions (in rough order of importance):

  1. test bench for all parts of our language technology efforts
  2. migrate to svn
  3. merge gt, kt and st into one, probably after the svn move
  4. more modularised make / build infra (prepare for smn, sms, sjd, others)
  5. close certain parts of the code repository (requires svn)
  6. set up the Leopard Server features for collaborative support:
    1. permanent chat rooms
    2. stored (and indexed) chat transcripts of the chat rooms
    3. iCal server / group calendars
    4. wiki
  7. wiki? (is part of Leopard Server) or other web-based documentation
  8. improve Forrest stability and i18n support
  9. reorganise the documentation content:
    1. differ between target groups
    2. get better grouping
    3. decide what to write in forrest and what in wiki (cf. Apertium for a similar split)
    4. update/add missing parts
  10. migrate lexicons to XML, splitting the task
    1. Name lexica (the Name project)
    2. Dictionaries (already in XML, task is to integrate them)
    3. Open POSes (Komi as a test case)
  11. change the look of the documentation web
  12. sfst? Both as replacement for xfst and for hunspell/open-source proofing tools
  13. investigate the NSIS installer, potentially replacing the InstallShield package from Polderland
  14. corpus content moved to Max Planck repositories?

TODO:

  • add CG regression test (Lene, Sjur, Trond)
  • make a test-all target that runs all tests we have (Børre, Sjur, Trond)
  • define and document testing routines (Børre, Sjur, Trond)
  • add Jabber account in iChat
    • check that all accounts are ready for iChat on the G5 (Børre)
    • UiT: Lene, Kimme, Saara (Trond)
  • prepare migration to svn (Børre, Sjur, Trond)
    • svn+ssh access is working, please test! svn checkout svn+ssh: //gtsvn.uit.no/usr/local/svn/repos/ gt
  • try to repair G5 accounts for iCal Server (Børre)

Linguistics

North Sámi

(nothing new, see proofing bugs below)

Lule Sámi

(nothing new, see proofing bugs below)

TODO:

  • sme->smj lexicon conversion to build bilingual lexicon resources, and increase smj coverage (Trond, Svenne).
  • Add the words when all words are ready.

South Sámi

Nothing new since last week.

Name lexicon infrastructure

Much work last week on the SD-terms collection, and the editing code relating to it. All old legacy code was cleaned and moved into the proper place, and deletion of entries was added.

Next up: to finish the work on the SD-terms collection (editing), then look at the name lexicon, and also dictionary search interface.

TODO:

  1. fix i18n bug in risten.no/G5 (so they will work without the proper locale request) (Sjur)
  2. fix bugs in lexc2xml; add comments to the log element (Saara)
  3. finish first version of the editing (Sjur)
  4. test editing of the xml files. If ok, then: ( Sjur, Thomas, Trond)
  5. make terms-smX.xml <=== automatically from propernoun-sme-lex.xml (add nob as well) (the morphological section should be kept intact, in e.g. propernoun-sme-morph.txt) (Sjur, Saara)
  6. convert propernoun-($lang)-lex.txt to a derived file from common xml files ( Sjur, Tomi, Saara)
  7. implement data synchronisation between risten.no and the cvs repo, and possibly other servers (ie the G5 as an alternative server to the public risten.no - it might be faster and better suited than the official one; also local installations could be treated the same way)
  8. start to use the xml file as source file
  9. clean terms-sme.xml such that all names have the correct tag for their use (e.g. @type=secondary) (Thomas, Maaren, linguists)
  10. merge placenames which are errouneously in different entries: e.g. Helsinki, Helsingfors, Helsset (linguists)
  11. publish the name lexicon on risten.no (Sjur)
  12. add missing parallel names for placenames (linguists)
  13. add informative links between first names like Niillas and Nils ( linguists)

Proofing tools

Hunspell

Printed out a partial fullword list, containing both the > and the » marks. Found out that double suffix stripping means that one continuation lexicon can point one lexicon further, but not more than that (sic!).

Made a python script that parses fullword lists, and prints out working .dic and .aff files. It has the same function as Polderlands program that parses our output. This script is slow, and has to be rewritten in C++ or Java.

TODO:

  1. review hunspell lexicon branch, merge with trunk if ok (Børre, Thomas)
  2. test latest lexicon (Sjur)
  3. add smj to the soup, make sure it works roughly as good as sme ( Børre, Thomas, Per-Eric)
    1. added to derivations, needs to be tested
  4. fix the remaining conversion bugs for sme ( Børre, Tomi)
  5. return to smj, and fix whatever is left to fix (Børre, Tomi)
  6. release a public beta at the end of April (Børre, Sjur)

Testing

Spelling Error Markup

TODO:

  • Set up ways of adding meta-information (source info, used in testing or not, added to lexicon or not) (Saara)
  • test new and nested error markup (Sjur)

Speller bugs

List of bugs returned from Polderland: 621, 630, 652, 656, 676.

Open issues based on test results :

sme

Version: Davvisámi, version 1.0.1, 2008-04-01

  • 425 - REGRESSION: other words from Divvun.no - two words rejected
  • 426 - comp words from Divvun.no - guoktedássásaš accepted - still OPEN
  • 435 - roman numbers - REGRESSION: inflection of single letter numbers rejected (but is ok in smj)
    • we should pregenerate all numbers once and for all, and store them in a separate lexicon file
  • 452 - REGRESSION: several lexical bugs - oažžuin + ožžuin
  • 595 - prefix+name wihtout hyphen (ovdaLot instead of ovda-Lot) - still OPEN
  • 600 - gen+hyph compound sámi-dáru - still OPEN
  • 603 - suomabealdi accepted - still OPEN
  • 606 - speller accepts VUOHTA compound - still OPEN
  • 607 - acro + hyphen - still OPEN
    • NRKGA is acro + clitic accepted without colon - what is correct?
  • 611 - double hyphen sugg still accepted - still OPEN
  • 613 - short gen. as second compound part - still OPEN
  • 619 - numerals and pronouns to NAMÁK and SASJ fails - still OPEN
  • 627 - prefix + hyhpen does not get accepted - still OPEN
  • 629 - a taking part in compounding without hyphen - still OPEN
  • 634 - PropGen+hyph+PropGen - still OPEN
  • 641 - numeral+noun compounds - still OPEN
  • 642 - noun/adj/proper + hyphen + ain - still OPEN
  • 644 - cased numeral+numeral compund - still OPEN
  • 646 - adverb + hyphen + noun - still OPEN
  • 647 - numerals+NOUN - still OPEN
  • 648 - unmotivated suggestions with numeral+noun - still OPEN
  • 649 - name + adj compound without hyphen - still OPEN
  • 654 - speller does not recognize ordinals on -nuppelogát - still OPEN
  • 655 - pron + nai - still OPEN
  • 658 - Suggestion saame - still OPEN

smj

Version: Julevsáme, version 1.0.1, 2008-04-01

  • 435 - roman number - single letter numbers now recognised
    • we should pregenerate all numbers once and for all, and store them in a separate lexicon file
    • please note that inflection of single letter numerals is fine in smj, as opposed to sme
  • 595 - prefix+name wihtout hyphen (tsåhkeLot instead of tsåhke-Lot) - still OPEN
  • 600 - gen+hyph compound sáme-dáro - still OPEN
  • 607 - acro + hyphen
    • NRKGA is acro + clitic accepted without colon - what is correct?
  • 616 - Bispadime-me-ráden - still OPEN, try to find an acro or abbr me
  • 619 - numerals and pronouns to NAMÁK and SASJ fails - still OPEN
  • 629 - a taking part in compound - still OPEN
  • 634 - rop gen + hyphen + Prop gen - still OPEN
  • 641 - numeral+noun compounds - still OPEN
  • 644 - cased numeral+numeral compund - still OPEN
  • 647 - numerals+NOUN - still OPEN
  • 648 - unmotivated suggestions with numeral+noun - still OPEN
  • 649 - name + adj compound without hyphen - still OPEN
  • 650 - noun prefix+name compound without hyphen - still OPEN
  • 658 - Suggestion saame - still OPEN

TODO:

  • compile new speller lexicons (Tomi)
  • document how compounding is controlled in the PLX conversion (Tomi)

Hyphenator bugs

Open issues based on test results :

sme

Lexicon version: Davvisámi, version 1.0.1, 2008-04-01

  • 468 - REGRESSION: Márkomeanu
  • 547 - REGRESSION: hyphen in front of vowel: Lotnolasealáhusas
  • 548 - REGRESSION: mid syllable hyphenation: Háliidivččen
  • 549 - REGRESSION: division without hyph: Váccedettiin
  • 673 - adj-derivations: guovttenuppelotčoarvvagiin (the word is not rec.)
  • 677 - NEW: Wrongly hyphenated ending -danidja - invalid

smj

Lexicon version: Julevsáme, version 1.0.1, 2008-04-01

  • 545 - REGRESSION: bad hyphenation in compounds: åhpadusorganisásjåvnån (not recognised)
  • 546 - REGRESSION: obligatory hyph rules seem to work in facultative manner: organisásjåvnån (not recognised)
  • 547 - REGRESSION: hyphen in front of vowel: Jienastimnjuolgadusá and Orgánajs

InDesign tools

We're waiting for an update from Polderland.

Releases

TODO:

  • update the Changes document (Sjur)
  • InDesign documentation (Sjur)
    • Norwegian translation received from Davvi Girji
  • public hunspell beta during first week of May - depends on the hunspell devel.
  • public 1.1 update of the Polderland-based tools during May

Other

Corpus contracts + open source

Now decided to wait until we have changed from cvs to svn.

TODO:

  • publish corpus contracts and project infra as open-source on NoDaLi-sta ( Sjur)

Travel to Røros and Östersund

Done, results somewhat mixed, but overall it was quite useful.

Next meeting, closing

The next meeting is 5.5.2008.

The meeting was closed at 11: 10.

Appendix - task lists for the next five days

Boerre

  • Hunspell lexicon conversion
  • prepare migration to svn (with Sjur, Trond)
  • release hunspell public beta at the end of April (with Sjur)
  • review hunspell lexicon branch (with Thomas), merge with trunk if ok
  • try to repair G5 accounts for iCal Server
  • make a test-all target that runs all tests we have
  • define and document testing routines
  • fix bugs!

Jovsset

  • follow up on sma corpus texts

Lene

  • get the ped content ready
  • Work on test routine with Trond and Sjur

Maaren

  • Put the list of possible sma corpus sources into a document

Per-Eric

  • try to find other authors who have smj texts digitaly.
  • Work with missing list from texts written by Sigga Tuolja Sandström.
  • Work with missing list same_dutkama_pgr.txt
  • Work with missing list sameriekta_tjoahkkagæsos.txt
  • Keep the contact with Ulf-Stefan Winka.
  • fix bugs!

Saara

  • add new XSL/XML headers for proofing test docs
  • Set up ways of adding meta-information for proofing correct corpus docs (source info, used in testing or not, added to lexicon or not)
  • implement the ped UI and functionality

Sjur

  • follow up on sma corpus texts
  • name db/risten.no
  • make an improved sma project plan
  • publish corpus contracts and project infra as open-source on NoDaLi-sta
  • prepare migration to svn (with Børre, Trond)
  • release hunspell public beta at the end of April (with Børre)
  • update the Changes document
  • follow-up on some Polderland-related bugs: 621, 630, 652, 656
  • test latest hunspell lexicon
  • InDesign documentation
  • send calendar info to Jovsset
  • add CG regression test with Lene and Trond
  • make a test-all target that runs all tests we have
  • define and document testing routines
  • fix bugs!

Thomas

  • look at test cases still not behaving properly
  • review hunspell lexicon branch with Børre
  • fix bugs!

Tomi

  • Hunspell lexicon conversion
  • document how compounding is controlled in the PLX conversion
  • fix double hyphen bugs
  • Make a pedagogical speller (after MA thesis is delivered)
  • fix bugs!

Trond

  • Help Jovsset with vislcg3 and sma
  • Set up Jabber for Lene, Kimme, Saara
  • Prepare svn migration (with Sjur, Børre)
  • Work on test routine with Lene and Sjur
  • make a test-all target that runs all tests we have
  • define and document testing routines
  • fix bugs!.