Belarusian and Russian/Work plan

Weekly plan

Semana	Dates	Coverage	Testvoc	Eval.		(%) cov. raw		(%) cov. trimmed		(%) WER		Bidix	Err.		Achieved ?
						bel	rus	bel→rus	rus→bel	bel→rus	rus→bel		bel→rus	rus→bel
1	18 abril—24 abril	55%				69.4	89.5	55.3	56.6	28.81	12.43	11,127	551	25,499	✓
2	25 abril—1 mayo	57.5%				70.9	89.5	65.6	65.7			42,225	3,691	76,627	✓
3	2 mayo—8 mayo	60%	pr, cnj*, adv	500		72.3	89.5	68.1	67.8	59.19	55.63	42,716	2,801	75,164
4	9 mayo—15 mayo	62.5%				72.6	89.7	69.8	68.7			42,884	5,505	76,994
5	16 mayo—22 mayo	67.5%				75.1	89.7	73.0	71.1			43,089	6,700	77,084
6	23 mayo—29 mayo	72.5%	prn, det			75.5		73.3	72.4			43,089	4,378	76,036	✓
7	30 mayo—5 junio	77.5%				80.6
8	6 junio—12 junio	82.5%
9	13 junio—19 junio	82.5%
10	20 junio—26 junio	82.5%
11	27 junio—3 julio	85%	n	500
12	4 julio—10 julio	86%
13	11 junio—17 julio	87%	vblex
14	18 julio—24 julio	87%
13	25 junio—31 julio	88%	adj
14	1 agosto—7 agosto	89%
15	8 agosto—14 agosto	90%		2000
16	15 agosto—21 agosto	90%

Calculating numbers

Errors (calculate in apertium-bel-rus)

$ sh dev/testvoc/generation.sh bel-rus | wc -l 
$ sh dev/testvoc/generation.sh rus-bel | wc -l

Bidix (calculate in apertium-bel-rus)

$ cat apertium-bel-rus.bel-rus.dix | grep '<l' | wc -l

Trimmed coverage (calculate in apertium-bel-rus)

$ cat bel.crp.txt | apertium -d . bel-rus-morph | sed 's/\$\W*\^/$\n^/g' > /tmp/bel.trim.coverage.txt
$ calc `cat /tmp/bel.trim.coverage.txt | grep -v '\*' | wc -l `/`cat /tmp/bel.trim.coverage.txt | wc -l`

Raw coverage (calculate in apertium-bel, apertium-rus)

$ cat bel.crp.txt | apertium -d . bel-morph | sed 's/\$\W*\^/$\n^/g' > /tmp/bel.raw.coverage.txt
$ calc `cat /tmp/bel.raw.coverage.txt | grep -v '\*' | wc -l `/`cat /tmp/bel.raw.coverage.txt | wc -l`