The university of Tromsø > Giellatekno
 

Meeting_2007-05-21

Meeting setup

  • Date: 21.05.2007
  • Time: 09.30 Norw. time
  • Place: Internet
  • Tools: SubEthaEdit, iChat

Agenda

  1. Opening, agenda review
  2. Reviewing the task list from last week
  3. Documentation - divvun.no
  4. Corpus gathering
  5. Corpus infrastructure
  6. Infrastructure
  7. Linguistics
  8. name lexicon infrastructure
  9. Spellers
  10. Other issues
  11. Summary, task lists
  12. Closing

1. Opening, agenda review, participants

Opened at 12: 50.

Present: Børre, Maaren, Per-Eric, Sjur, Steinar, Thomas, Tomi, Trond

Absent: Saara

Agenda accepted as is.

2. Updated task status since last meeting

Børre

  • add sma texts to the corpus repository
    • not done
  • collect a list of PR recipients, forward to Berit Karen Paulsen
    • not done
  • run all known spelling errors in the prooftest corpus through the speller
    • not done
  • add extraction of all known spelling errors in the regular corpus (not the prooftest corpus), and check that they are properly corrected
    • not done
  • update installer packages with latest speller lexicon
    • not done
  • update and fix our documentation and infrastructure as Steinar finds problem areas - low priority
    • not done
  • study the Hunspell formalism in detail
    • nothing new
  • test the typos.txt list, and check that all entries are properly corrected
    • test done, not added to cvs
  • add info to front page (incl. download links)
    • not done
  • write separate page with detailed info (incl. download links)
    • a separate page for the beta speller, with installation instructions, etc.
      • began working on it
  • translate press release, web pages
    • not done
  • follow up the server OS on victorio
    • nothing new
  • contact Davvi Girji / Mikal Aase
    • not done
  • fix bugs!
    • number 396 fixed
  • other
    • built spellers

Maaren

  • lexicalise actio compounds
  • Manually mark speller test documents for typos

Per-Eric

  • expand the smj typos list
    • continued the work with typos list
  • add missing smj words
    • continued the work with searching missing words in smj
  • research the –k ending, whether it is a clitic or a regular inflection
    • working on it

Saara

  • improve cgi-bin scripts
    • add new features to the paradigm generator
    • paradigm generator for Lule Sámi
  • add new XSL/XML headers for proofing test docs
  • compilation of verb lists
  • read the manual for graphical corpus interface and try to add files with Lars.
  • fix bugs!

Sjur

  • finish press release for the beta
    • not done
  • collect a list of PR recipients
    • Trond has a list he has used earlier - looks like a good starting point
  • run all known spelling errors in the corpus through the speller
    • not done
  • document the AppleScript testing tool
    • not finished
  • integrate regression self tests with the make file
    • not done
  • make improved smj speller
    • done - Børre did it
  • improve speller test bench
    • not done
  • update installer packages with latest speller lexicon
    • did it once, sent it to Polderland for technical testing, needs to be updated with the latest speller lexicons. Also: making the installer needs to be integrated in a make file.
  • integrate the ccat speller testing options in the make file
    • not done
  • fix internet setup for Per-Eric's satelite modem
    • nothing more - waiting for answer
  • ask for Linux version of the Polderland command-line speller (for victorio)
    • done - waiting for quote
  • ask for mklex for Linux (victorio) from Polderland (for victorio)
    • done - waiting for quote
  • look over the Bugzilla status mails
    • not done
  • contact Davvi Girji / Mikal Aase
    • not done
  • ask Xerox for a commercial lisense for the xfst tools on the G5
    • not done
  • fix bugs!
  • other:
    • helped with integrating compound restriction tags
    • had a meeting with SD and Davvi Girji about an electronic dictionary project of theirs; also briefly discussed corpus material from them.

Steinar

  • Beta testing: Align manually (shorter texts)
  • Manually mark speller test texts for typos (making them into gold standards), add the texts to the orig/catalogue
    • added some facta texts
  • Complete the semantic sets in sme-dis.rle
    • no work this week
  • missing lists
    • no work this week
  • Look at the actio compound issue when adding from missing lists
    • not done
  • fix bugs!

Thomas

  • work with compounding
    • worked hard
  • Lack of lowering before hyphen: Twol rewrite.
    • not done
  • translate beta release docs to sme and smj
    • not done
  • Add potential speller test texts
    • not done
  • research the –k ending, whether it is a clitic or a regular inflection
    • done
  • fix bugs!
    • worked some

Tomi

  • add compound restrictions to both spellers
    • Saara wrote a script to include the tags in the lexc source
  • make PLX conversion test sample; add conversion testing to the make file
    • not done
  • improve prefix and middle-noun PLX conversion
    • not done
  • integrate the ccat speller testing options in the Makefile
    • not done
  • fix bugs!
    • fixed speller beta bugs
      • improved compound flagging
      • improved numeral conversion

Trond

  • Work on the parallel corpus issues
    • Worked on monolingual texts a bit.
  • Add Steinars corr texts to the relevant catalogue
    • Not done
  • Postpone these tasks to after the beta:
    • update the smj proper noun lexicon, and refine the morphological analysis, cf. the propernoun-smj-lex.txt
    • Go through the Num bugs
  • collect a list of PR recipients
    • Done
  • check with Lars when more aligned texts are ready
    • Saara talks to Lars.
  • fix bugs!.
    • Fixed some, added more.

3. Documentation

The open documentation issues fall into these three categories:

  • Beta documentation for testers
  • Documentation for the online corpora
  • General documentation improvement after Steinar's test (for open-source release)

TODO:

  • write form to request corpus user account (Børre, Sjur, Trond)
    • delayed till after the beta release
  • document how to apply for access to closed corpus, and details on the corpus and its use in general (Børre, Sjur, Trond)
    • delayed till after the beta release
  • correct and improve it based on feedback from Steinar ( Børre)
    • low priority
  • beta documentation (see separate beta section below)

4. Corpus gathering

TODO:

  • sme texts: no new additions, fix corpus errors during this month ( Børre, Trond, Saara)
  • missing nob parallel texts should be added if such holes are found ( Børre, Trond)
  • Go through the list of missing or errouneous nob texts, based upon Saara's perfect list (Børre, Trond)
  • add sma texts to the corpus repository (Børre)
  • contact Davvi Girji / Mikal Aase ( Børre, Sjur)

5. Corpus infrastructure

Access to licensed texts

We have four novels in our corpus, three of them will soon be added to the web interface. Could they be added using the present, free-access web interface (which only gives access to hit words with context, at most three sentences), or should we have a password-protected version of the web interface to allow access to these documents?

The present limitation of maximum three sentences at a time should be enough to stop all but the most dedicated attempts at misuse - and such dedicated attempts can not be stopped anyway, but should be handed over to the police if they happen. Thus, it should be ok to allow search access to the novels using the present, open web interface, ie without a user name and password.

Update: the web interface has been changed to always require registration before use, and usage requires a user name and password, thus all use can be logged, and mis-use can be better traced. The only thing required to get a username is to fill in a simple web form, accessible at http://omilia.uio.no/swamp/index.php?main=sami. The corpus itself is, as before, available at http://omilia.uio.no/glossa/html/index_dev.php?corpus=samno.

Alignment

TODO

  • check with Lars when more texts are ready (Trond)

6. Infrastructure

TODO:

  • update and fix our documentation and infrastructure as Steinar finds problem areas (Børre)
  • fix internet setup for Per-Eric's satelite modem (Sjur, Børre)
    • this influences iChat, SEE sharing, and ARD connetions
      • did some investigation last week, waiting for an answer from the modem maker
  • Add a Lule Sámi paradigm generators, and link the p-smj.en.html page to it. ( Saara)
    • done
  • Translate the paradigm pages to Sámi, fix all the se links ( Børre, Trond)
    • we have separate documents, but they are not translated, and not linked
      • Saara has updated much of the infra for this, with better support for all the tools, and properly localised pages.

7. Linguistics

North Sámi

TODO:

  • lexicalise actio compounds. Example: vuolggasadji vs. vuolginsadji ( Maaren)
    • possibly turn on free compounding as part of the PLX conversions (ie free compounding in the spellers, but not in the analyzers/transducers)
  • fix stuorra-oslolaš lower case o ( Sjur, Thomas, Trond)
    • postponed till after the public beta

Lule Sámi

TODO:

  • refine smj proper noun lexica, cf. the propernoun-smj-lex.txt ( Thomas, Trond)
  • research the –k ending, whether it is a clitic or a regular inflection, and where it applies (Per-Eric, Thomas)
    • it is a clitic, an allomorphic variant of -ge (after vowels). The variation should be handled by two-level rules. The clitic thus needs to be generated as part of the regular PLX conversion (for spellers).
  • write two-level rules to handle -ge/-k clitic alternation (Thomas)

8. Name lexicon infrastructure

Decisions made in Tromsø can be found in this meeting memo.

TODO:

  1. fix bugs in lexc2xml; add comments to the log element (Saara)
  2. finish first version of the editing (Sjur)
  3. test editing of the xml files. If ok, then: ( Sjur, Thomas, Trond)
  4. 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)
  5. convert propernoun-($lang)-lex.txt to a derived file from common xml files ( Sjur, Tomi, Saara)
  6. 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)
  7. start to use the xml file as source file
  8. clean terms-sme.xml such that all names have the correct tag for their use (e.g. @type=secondary) (Thomas, Maaren, linguists)
  9. merge placenames which are errouneously in different entries: e.g. Helsinki, Helsingfors, Helsset (linguists)
  10. publish the name lexicon on risten.no (Sjur)
  11. add missing parallel names for placenames (linguists)
  12. add informative links between first names like Niillas and Nils ( linguists)

9. Spellers

OOo speller(s)

TODO after the MS Office Beta is delivered:

  • add Hunspell data generation to the lexc2xspell (Tomi - after the PLX data generation is finished)
  • study the Hunspell formalism in detail (Børre, Sjur, Tomi)

Testing

Spelling Error Markup

TODO:

  • Manually mark test texts for typos (making them into gold standards) ( Steinar)
  • Set up ways of adding meta-information (source info, used in testing or not, added to lexicon or not)
  • Conduct tests on new beta versions on the basis of the unspoiled gold standard documents (whoever has time), and fill in data from testing (the testers: who?)
  • include the files already publically tested into the prooftest cataloge ( Steinar, Trond)
  • test each version before beta release

Testing tools

TODO:

  • document the AppleScript testing tool (Sjur)
  • improve speller test bench (Sjur)
    • integrate the ccat speller testing options in the Makefile (Sjur, Tomi)
  • ask for Linux version of the Polderland command-line speller (for victorio) ( Sjur)
    • done, waiting for a price quote (it is outside the present contract)

Regression tests

TODO:

  • add extraction of all known spelling errors in the corpus (not the prooftest corpus), and check that they are properly corrected ( Børre, Sjur)
  • test the typos.txt list, and check that all entries are properly corrected ( Børre, Sjur)
  • consider how to do a regression self-test, ie, how to test the full wordlist (Børre, Sjur)
    • extract all the base forms in the lexicon, and run them through the speller
    • extract all SUB-marked entries, and run them through the lexicon
      • integrate these in the make file (Sjur)

Localisation

We need to translate the info added to our front page (and a separate page) regarding the beta release. Also the press release needs to be translated.

TODO:

  • translate beta release docs to sme ( Thomas)
  • translate beta release docs to smj ( Thomas)

Lexicon conversion to the PLX format

It is now going pretty well, but we still need to transfer the files from victorio to a Mac with Polderlands mklexSami* tools. Børre is regularly updating the spellers.

TODO:

  • install larger disks, new RAM on the G5 when they arrive (Børre)
  • ask for mklex for Linux (victorio) from Polderland (Sjur)
    • waiting for a price quote.
  • ask Xerox for a commercial lisense for the xfst tools on the G5 (Sjur)
    • not done
  • add compounding restrictions to the PLX convertion (Tomi)

Compounding restrictions

How to include compounding restriction comment tags in the transducers:

giv0ri:giv'ri ALBMI ; !+SgNomCmp +SgGenCmp +PlGenCmp
=> (using a perl script or similar)
+SgNomCmp+SgGenCmp+PlGenCmpgiv0ri:giv'ri ALBMI ; !+SgNomCmp +SgGenCmp +PlGenCmp

TODO:

  1. improve prefix conversion to PLX (Tomi)
  2. improve middle noun conversion to PLX (Tomi)
  3. improve noun + adjective PLX conversion: ( Tomi)
    1. compounding stems - how do we generate them? Using the java client? +SgNomCmp+Cmpnd = sáme–, should give the correct compounding stem, shouldn't it? We want to optionally go from: sáme- NLI to sáme NL: - NLI (->) NL, which means we should be able to extract correct compounding stems using xfst methods only.
    2. compounding tags - we need to obey them when making the transducers. Suggestion - see above.
  4. make conversion test sample; add conversion testing to the make file ( Tomi)
    1. to regression test / QA the PLX conversion.
  5. improve number conversion (Børre, Tomi)
  6. ask for larger disk for the web server (Trond, Børre)

Errors in latest speller build

Build: Davvisámi, PBeta 1, 2007-05-19

  • first part of multiword expressions not accepted
  • errors in numbers - fixed
  • no clitics on nouns, adjs
  • compound restrictions/rules not yet integrated in speller fsts (but the tags necessary to do this can now be included in the transducer)
  • suomasápmelaš vs suopmasápmelaš - now both are accepted (which is an improvement - earlier only suopmasápmelaš was accepted). What is correct? Both?
  • smj: öä not accepted, only øæ (except for lexicalised names).
  • ŋ and ñ and ń in smj

Example: maŋŋel vs maññel / mańńel

  • maññel - ok
  • mańńel - ń is not recognised as part of the word, should be reported to Polderland

Try TODO before final public beta:

  • fix clitics (Tomi)
  • include improved numbers (Børre)

TODO after the final public beta:

  • first part of multiword expressions not accepted (Tomi)
  • compound restrictions/rules not yet integrated in speller fsts (but the tags necessary to do this can now be included in the transducer) (Tomi)
  • suomasápmelaš vs suopmasápmelaš - now both are accepted (which is an improvement - earlier only suopmasápmelaš was accepted). What is correct? Only the first one. The second one should be banned by the compound restrictions.
  • smj: öä not accepted, only øæ (except for lexicalised names) ( Thomas)
  • ŋ and ñ and ń in smj ( Sjur, Polderland)

Public Beta release

Schedule: May 22, evening: deadline for linguistic fixes and updates May 23, morning: Release Candidate (RC) spellers ready for testing May 23: press release, untranslated web pages README May 24: Linguistic testing finished May 25: translation of press release web pages translated README translated

DEADLINE/RELEASE: Tuesday May 29.

TEST before public beta release:

  • Actio compounds (Thomas, Steinar)
  • download and installation on all platforms, following our instructions ( Steinar, Børre)
  • deinstallation (Steinar, Børre)
  • linguistic testing (typos, at least one of the documents marked up by Steinar) (Thomas, Børre)

TODO:

  • working and updated smj speller (Sjur, Tomi)
    • done
  • finish press release (Sjur)
  • add info to front page (incl. download links) (Børre)
    • download page made, only needs to add the speller beta when it is ready.
  • write separate page with detailed info (incl. download links) (Børre)
    • a separate page for the beta speller, with installation instructions, etc.
  • translate press release, web pages (Børre, Thomas, whoever)
  • forward list of PR recipients to Berit Karen Paulsen ( Sjur)
    • update: the persons to contact are: Pål Hivand or John Markus Kuhmunen
  • update installer packages with latest speller lexicons (Børre (Win), Sjur (Mac))
  • README file: ( Børre, Sjur)
    • list of known issues
    • how to give us feedback
    • what to give us feedback about (including: send in your user dictionary after a month or so)
    • inform about the rights to the source, as well as the availability of the speller
    • needs to be written in all languages
    • make a PDF of the readme, to avoid encoding problems; text versions should be included too

10. Other

Summer vacation

When are we taking it? Please fill in the table below:

Name Starting Ending
Børre x|x
Maaren x|x
Per-Eric x|x
Saara x|x
Sjur x|x
Steinar x|x
Thomas 9.7. 12.8.
Tomi 9.7. 5.8.
Trond 25.6. |x

Divvun people also need to send the dates to Julie Eira.

Corpus contracts

TODO:

  • publish corpus contracts and project infra as open-source on NoDaLi-sta ( Sjur)
    • delayed until the public beta is out the door

Bug fixing

41 open Divvun/Disamb bugs, and 23 risten.no bugs

TODO:

  • look over the Bugzilla status mails (Børre)

11. Next meeting, closing

The next meeting is 29.5.2007, 10: 30 Norwegian time.

The meeting was closed at 11: 56.

Appendix - task lists for the next week

Boerre

iCal

  • add sma texts to the corpus repository
  • run all known spelling errors in the prooftest corpus through the speller
  • add extraction of all known spelling errors in the regular corpus (not the prooftest corpus), and check that they are properly corrected
  • update and fix our documentation and infrastructure as Steinar finds problem areas - low priority
  • study the Hunspell formalism in detail - after beta release
  • test the typos.txt list, and check that all entries are properly corrected
  • add info to front page (incl. download links)
  • write separate page with detailed info (incl. download links)
    • a separate page for the beta speller, with installation instructions, etc.
  • translate press release, web pages
  • contact Davvi Girji / Mikal Aase
  • include improved numbers in beta release speller
  • test download and installation of beta release candidate on all platforms
  • test deinstallation of beta release candidate on all platforms
  • linguistic testing of speller beta release candidate
  • update Win installer package with latest speller lexicons
  • write and prepare README file
  • fix bugs!

Maaren

iCal

  • lexicalise actio compounds
  • Manually mark speller test documents for typos

Per-Eric

iCal

  • expand the smj typos list
  • add missing smj words

Saara

iCal

  • improve cgi-bin scripts
    • add new features to the paradigm generator
    • paradigm generator for Lule Sámi
  • add new XSL/XML headers for proofing test docs
  • compilation of verb lists
  • read the manual for graphical corpus interface and try to add files with Lars.
  • fix bugs!

Sjur

iCal

  • finish press release for the beta
  • run all known spelling errors in the corpus through the speller
  • document the AppleScript testing tool
  • integrate regression self tests with the make file
  • improve speller test bench
  • integrate the ccat speller testing options in the make file
  • fix internet setup for Per-Eric's satelite modem
  • look over the Bugzilla status mails
  • contact Davvi Girji / Mikal Aase
  • ask Xerox for a commercial lisense for the xfst tools on the G5
  • check with Sámi publishing houses whether support for CS2 is still needed
  • send Beta RC Win and Mac installers to Polderland for testing
  • ń not considered word char in smj, not corrected to ŋ
  • forward list of PR recipients to SD
  • update Mac installer package with latest speller lexicons
  • write and prepare README file
  • fix bugs!

Steinar

iCal

  • test actio compounds before public beta release
  • test download and installation on all platforms, following our instructions
  • test deinstallation
  • Beta testing: Align manually (shorter texts)
  • Manually mark speller test texts for typos (making them into gold standards), add the texts to the orig/catalogue
  • Complete the semantic sets in sme-dis.rle
  • missing lists
  • fix bugs!

Thomas

iCal

  • work with compounding
  • Lack of lowering before hyphen: Twol rewrite.
  • translate beta release docs to sme and smj
  • implement –k ending in twol
  • Test Actio compounds in speller beta release candidate
  • linguistic testing of speller beta release candidate
  • smj: öä not accepted, only øæ (except for lexicalised names)
  • fix bugs!

Tomi

iCal

  • add compounding restrictions to the PLX convertion
  • make PLX conversion test sample; add conversion testing to the make file
  • improve prefix and middle-noun PLX conversion
  • integrate the ccat speller testing options in the Makefile
  • fix clitics in beta release sme speller
  • first part of multiword expressions not accepted
  • fix bugs!

Trond

iCal

  • Work on the web corpus issues
  • Postpone these tasks to after the beta:
    • update the smj proper noun lexicon, and refine the morphological analysis, cf. the propernoun-smj-lex.txt
    • Go through the Num bugs
  • fix bugs!.