GramDivvun-evaluerenčoahkkin

10.06.2020
10:15-

Børre, Linda

doing:
FileMerge: comparing report.correct-no-gs-20200609.txt and      report.correct-no-gs.txt

problems:
duoddjái is fixed on commandline, but not in the evaluation (Ávnnas válljema rájes gitta gárvves duoddjái.)
old se.zcheck -≥ doing make in lang-sme should be the cure for this

ls -l $GTLANGS/lang-sme/tools/grammarcheckers/se.zcheck
-rw-r--r--  1 lwi001  1907360568  54323268  5 Jun 12:44 /Users/lwi001/lang-sme/tools/grammarcheckers/se.zcheck

todo:
* check the date of the current se.zcheck file: ls -l $GTLANGS/lang-sme/tools/grammarcheckers/se.zcheck
* lang-sme: run make
* check if there is a new se.zcheck file: ls -l $GTLANGS/lang-sme/tools/grammarcheckers/se.zcheck
* run again:
{{{
rm -rf $GTFREE/correct-no-gs/converted # sihkku boares konverterejuvvon fiillaid
svn up $GTFREE/orig/sme/ # viežžá ođasmahttojuvvon fiillaid
convert2xml --goldstandard $GTFREE/orig/sme/grammar-realword # konvertere fiillaid
cd $GTFREE
$GTCORE/devtools/gramcheck_tester2.py $GTLANGS/lang-sme/tools/grammarcheckers/se.zcheck $GTFREE/correct-no-gs/converted/sme # vuodjá grammarcheckera, ráhkada .pickle-fiilla
$GTCORE/devtools/gramcheck_comparator.py $GTLANGS/lang-sme/tools/grammarcheckers/se.zcheck correct-no-gs --filtererror errorlex errormorphsyn # lohká .pickle-fiilla, ráhkada raportta
}}}



check:
Son diđii ahte oktii ollii Johani Nousuniemi ealahat ahkkái, ja dalle lei sus vejolašvuohta beassat dan doibmii. --- ealahat
* 1691 gonagasreivve gohččui čuohppat dimbariid Várjjaga ođđa girkui ja Čáhcesullo ođđa ámtamánnigárdimii Njávdánvuvddiin (Niemi 1983: 333). -- typo
* Iđđes gohččui áddjá báhpa ja su eamida boahtit guossái.
* Dát dovdomearkkat leat maid boazodoalu muohtakarakteriseremis, mas omd. geavahit dajaldagaid assás geardni dan saddjái go dárkilis mihtidemiin, ahte lea omd. 10 cm geardni {Eira et al., 2011, sisa sáddejuvvon}.  --- et al.
* Árbbolaččat- teakstaduojis lea teaksta hábmejuvvon nu ahte muhtimin persovnna perspektiiva ihtá goalmmátpersovdna-muitaleaddjin, dan saddjái go ahte mun-muitaleaddji govvida persovnna jurdagiid.  <- /Users/lwi001/freecorpus/correct-no-gs/converted/sme/grammar-realword/real-nsgill-asgnom.correct.txt.xml
* Duogáš vuohkkái, ahte addit jahkásaš dieđáhusa Stuorradiggái Sámedikki doaimma birra, ja prinsihpalaš dieđáhusa Norgga sámepolitihka birra juohke stuorradiggeáigodagas, lea Sámi vuoigatvuođalávdegotti vuosttas oassečielggadus NAČ 1984:18 "Sámiid riektidili birra".  <- /Users/lwi001/freecorpus/correct-no-gs/converted/sme/grammar-realword/real-nsgill-asgnom.correct.txt.xml --- riektedili
* 30-vuođas daddjo, ahte maŋŋá go ekonomiija sirdašuvai iešbirgejumis márkanekonomiijii, de nissonolbmuid barggut eai atnon šat seamma árvosažžan boazodoalus go almmáiolbmuid barggut.  <- /Users/lwi001/freecorpus/correct-no-gs/converted/sme/grammar-realword/real-imprtsg1-derpassprfprc.correct.txt.xml
~~~~~~ seamma

* Giellaplánen juohkku lingvistihkas árbevirolaččat golbma dássái: stáhtus-, korpus- ja Robert L. Coopera (1989) modeallas maiddái giellaoahpahusplánemii (acquisititon planning).  <- /Users/lwi001/freecorpus/correct-no-gs/converted/sme/grammar-realword/real-imprtdu1-derpassprssg3.correct.txt.xml (other language) --- done
~~~~~~


Thank you:
B.G.
for providing the grammarchecker evaluator (a program for the automatic evaluation),
technical support and fruitful methodological and linguistic discussions.

The grammarchecker evaluator consists of 2 scripts for evaluating the current grammarchecker (or part of it) and counting false/true positives/negatives and providing statistics for precision/recall, both general and ordered by error type.
$GTCORE/devtools/gramcheck_tester2.py # makes the raw data
$GTCORE/devtools/gramcheck_comparator.py # makes a text based report from the raw data

Discuss with Tommi
	# we need a rule that excludes the selection of Spellrelax if there is any other reading that does not have Err/...