Talk:Afrikaans and English

Performance

Some basic statistics on the quality (or rather coverage) of the Afrikaans->English system.

• Unknown, * — these words have not been able to be analysed (they are not in the monolingual dictionary).
• Transfer error, # — there has been an error in transfer (some symbols cannot be resolved).
• Untranslated, @ — the word is analysed but a translation does not appear in the bilingual dictionary.

12 August 2007

All texts from Die Volksblad.

Filename	Word count	Unknown	(%)	Transfer error	(%)	Untranslated	(%)
001.txt	383	56	~14.62	48	~12.53	3	~0.783
002.txt	582	181	~31.09	42	~7.216	12	~2.061
003.txt	943	300	~31.81	97	~10.28	16	~1.696
004.txt	747	235	~31.45	44	~5.890	22	~2.945
005.txt	201	31	~15.42	13	~6.467	5	~2.487
006.txt	307	73	~23.77	20	~6.514	5	~1.628
007.txt	417	117	~28.05	38	~9.112	6	~1.438
008.txt	405	114	~28.14	21	~5.185	9	~2.222
009.txt	304	74	~24.34	35	~11.51	5	~1.644
010.txt	157	21	~13.37	16	~10.19	0	0
011.txt	335	66	~19.70	33	~9.850	3	~0.895
012.txt	464	80	~17.24	60	~12.93	4	~0.862
013.txt	155	29	~18.70	21	~13.54	1	~0.645
014.txt	423	69	~16.31	60	~14.18	5	~1.182
015.txt	410	110	~26.82	33	~8.048	11	~2.682
016.txt	346	83	~23.98	38	~10.98	11	~3.179
017.txt	357	98	~27.45	31	~8.683	9	~2.521
018.txt	491	74	~15.07	59	~12.01	10	~2.036
019.txt	378	84	~22.22	27	~7.142	6	~1.587
020.txt	592	135	~22.80	58	~9.797	4	~0.675

22 July 2007

All texts from Die Volksblad.

Filename	Word count	Unknown (%)	Transfer error (%)	Untranslated (%)
001.txt	383	59 (~15.40)	44 (~11.48)	2 (~0.522)
002.txt	582	217 (~37.28)	29 (~4.982)	4 (~0.687)
003.txt	943	364 (~38.60)	77 (~8.165)	4 (~0.424)
004.txt	747	285 (~38.15)	34 (~4.551)	11 (~1.472)
005.txt	201	33 (~16.41)	12 (~5.970)	5 (~2.487)
006.txt	307	87 (~28.33)	18 (~5.863)	0 (0)
007.txt	417	142 (~34.05)	23 (~5.515)	1 (~0.239)
008.txt	405	130 (~32.09)	11 (~2.716)	2 (~0.493)
009.txt	304	94 (~30.92)	28 (~9.210)	2 (~0.657)
010.txt	157	24 (~15.28)	16 (~10.19)	0 (0)
011.txt	335	75 (~22.38)	31 (~9.253)	1 (~0.298)
012.txt	464	83 (~17.88)	56 (~12.06)	4 (~0.862)
013.txt	155	30 (~19.35)	20 (~12.90)	0 (0)
014.txt	423	85 (~20.09)	60 (~14.18)	1 (~0.236)
015.txt	410	137 (~33.41)	22 (~5.365)	5 (~1.219)
016.txt	346	100 (~28.90)	36 (~10.40)	3 (~0.867)

08 July 2007

Newspaper	Filename	Total	Unknown	Percentage OOV
Die Volksblad	001.txt	383	110	28.7
Die Volksblad	002.txt	582	258	44.3
Die Volksblad	003.txt	943	433	45.9
Die Volksblad	004.txt	747	337	45.1
Die Volksblad	005.txt	201	71	35.3
Die Volksblad	006.txt	307	121	39.4
Die Volksblad	007.txt	417	185	44.3
Die Volksblad	008.txt	405	157	38.7
Die Volksblad	009.txt	304	120	39.4
Beeld	001.txt	309	116	37.5

26 June 2007

Over a selection of texts from Die Volksblad and Beeld:

Newspaper	Unknown/Total	Filename	Percentage OOV
Die Volksblad	117/383	001.txt	30.5
Die Volksblad	267/582	002.txt	45.8
Die Volksblad	449/943	003.txt	47.6
Die Volksblad	347/747	004.txt	46.4
Die Volksblad	77/201	005.txt	38.3
Beeld	123/309	001.txt	39.8