Skip to content

Commit

Permalink
Add FastSpell to accuracy reports (#188)
Browse files Browse the repository at this point in the history
  • Loading branch information
pemistahl committed Nov 28, 2023
1 parent a5c28f5 commit 3f1359c
Show file tree
Hide file tree
Showing 147 changed files with 2,972 additions and 185 deletions.
769 changes: 702 additions & 67 deletions README.md

Large diffs are not rendered by default.

152 changes: 76 additions & 76 deletions accuracy-reports/aggregated-accuracy-values.csv

Large diffs are not rendered by default.

16 changes: 16 additions & 0 deletions accuracy-reports/fastspell-aggr/Afrikaans.txt
Original file line number Diff line number Diff line change
@@ -0,0 +1,16 @@
##### Afrikaans #####

>>> Accuracy on average: 72.83%

>> Detection of 1000 single words (average length: 8 chars)
Accuracy: 49.90%
Erroneously classified as English: 16.40%, Dutch: 3.80%, French: 3.70%, Unknown: 3.70%, Italian: 2.30%, Swedish: 2.20%, Danish: 2.10%, German: 1.90%, Bokmal: 1.80%, Finnish: 1.50%, Polish: 1.50%, Turkish: 1.40%, Estonian: 1.20%, Indonesian: 1.00%, Portuguese: 0.90%, Spanish: 0.80%, Catalan: 0.50%, Azerbaijani: 0.40%, Hungarian: 0.40%, Lithuanian: 0.40%, Esperanto: 0.30%, Slovene: 0.30%, Latin: 0.20%, Romanian: 0.20%, Arabic: 0.10%, Belarusian: 0.10%, Chinese: 0.10%, Croatian: 0.10%, Czech: 0.10%, Nynorsk: 0.10%, Persian: 0.10%, Russian: 0.10%, Serbian: 0.10%, Slovak: 0.10%, Swahili: 0.10%, Welsh: 0.10%

>> Detection of 1000 word pairs (average length: 16 chars)
Accuracy: 73.60%
Erroneously classified as English: 10.10%, Danish: 2.40%, Dutch: 2.40%, French: 2.20%, Finnish: 1.30%, Italian: 1.20%, Unknown: 0.90%, Estonian: 0.80%, Polish: 0.70%, Bokmal: 0.60%, Swedish: 0.60%, Turkish: 0.50%, Azerbaijani: 0.30%, Indonesian: 0.30%, Portuguese: 0.30%, Esperanto: 0.20%, Hungarian: 0.20%, Romanian: 0.20%, Welsh: 0.20%, Catalan: 0.10%, Croatian: 0.10%, German: 0.10%, Latin: 0.10%, Latvian: 0.10%, Nynorsk: 0.10%, Serbian: 0.10%, Slovak: 0.10%, Slovene: 0.10%, Tamil: 0.10%

>> Detection of 1000 sentences (average length: 102 chars)
Accuracy: 95.00%
Erroneously classified as English: 2.00%, Unknown: 1.10%, French: 1.00%, Dutch: 0.20%, Italian: 0.20%, Azerbaijani: 0.10%, Bokmal: 0.10%, Danish: 0.10%, Estonian: 0.10%, Turkish: 0.10%

16 changes: 16 additions & 0 deletions accuracy-reports/fastspell-aggr/Albanian.txt
Original file line number Diff line number Diff line change
@@ -0,0 +1,16 @@
##### Albanian #####

>>> Accuracy on average: 66.10%

>> Detection of 1000 single words (average length: 8 chars)
Accuracy: 34.90%
Erroneously classified as English: 15.20%, Italian: 5.00%, German: 4.90%, Unknown: 4.80%, Turkish: 2.90%, Portuguese: 2.80%, French: 2.60%, Spanish: 2.50%, Swedish: 2.30%, Esperanto: 2.20%, Serbian: 2.10%, Dutch: 1.80%, Croatian: 1.70%, Slovene: 1.70%, Bokmal: 1.60%, Indonesian: 1.60%, Finnish: 1.40%, Czech: 0.90%, Polish: 0.90%, Romanian: 0.70%, Bosnian: 0.60%, Malay: 0.60%, Azerbaijani: 0.50%, Catalan: 0.50%, Vietnamese: 0.50%, Danish: 0.40%, Hungarian: 0.40%, Lithuanian: 0.30%, Afrikaans: 0.20%, Basque: 0.20%, Estonian: 0.20%, Nynorsk: 0.20%, Russian: 0.20%, Chinese: 0.10%, Icelandic: 0.10%, Persian: 0.10%, Slovak: 0.10%, Swahili: 0.10%, Tagalog: 0.10%, Urdu: 0.10%

>> Detection of 1000 word pairs (average length: 15 chars)
Accuracy: 65.50%
Erroneously classified as English: 8.00%, Serbian: 2.70%, Unknown: 2.10%, Esperanto: 1.80%, Italian: 1.80%, German: 1.50%, Portuguese: 1.40%, Croatian: 1.30%, Finnish: 1.30%, Slovene: 1.20%, Swedish: 1.20%, French: 1.10%, Polish: 1.00%, Turkish: 1.00%, Bokmal: 0.90%, Indonesian: 0.90%, Malay: 0.70%, Romanian: 0.70%, Spanish: 0.70%, Dutch: 0.60%, Estonian: 0.50%, Bosnian: 0.40%, Hungarian: 0.30%, Lithuanian: 0.30%, Vietnamese: 0.30%, Catalan: 0.20%, Danish: 0.20%, Afrikaans: 0.10%, Basque: 0.10%, Icelandic: 0.10%, Latin: 0.10%

>> Detection of 1000 sentences (average length: 118 chars)
Accuracy: 97.90%
Erroneously classified as Serbian: 0.40%, Croatian: 0.20%, English: 0.20%, Esperanto: 0.20%, Portuguese: 0.20%, Finnish: 0.10%, French: 0.10%, German: 0.10%, Malay: 0.10%, Spanish: 0.10%, Swedish: 0.10%, Turkish: 0.10%, Unknown: 0.10%, Urdu: 0.10%

16 changes: 16 additions & 0 deletions accuracy-reports/fastspell-aggr/Arabic.txt
Original file line number Diff line number Diff line change
@@ -0,0 +1,16 @@
##### Arabic #####

>>> Accuracy on average: 95.50%

>> Detection of 1000 single words (average length: 6 chars)
Accuracy: 89.20%
Erroneously classified as Persian: 5.20%, Unknown: 4.20%, Urdu: 0.90%, English: 0.40%, Ukrainian: 0.10%

>> Detection of 1000 word pairs (average length: 14 chars)
Accuracy: 97.60%
Erroneously classified as Unknown: 1.40%, Persian: 1.00%

>> Detection of 1000 sentences (average length: 89 chars)
Accuracy: 99.70%
Erroneously classified as Unknown: 0.30%

16 changes: 16 additions & 0 deletions accuracy-reports/fastspell-aggr/Armenian.txt
Original file line number Diff line number Diff line change
@@ -0,0 +1,16 @@
##### Armenian #####

>>> Accuracy on average: 100.00%

>> Detection of 1000 single words (average length: 9 chars)
Accuracy: 100.00%
Erroneously classified as

>> Detection of 1000 word pairs (average length: 18 chars)
Accuracy: 100.00%
Erroneously classified as

>> Detection of 1000 sentences (average length: 122 chars)
Accuracy: 100.00%
Erroneously classified as

16 changes: 16 additions & 0 deletions accuracy-reports/fastspell-aggr/Azerbaijani.txt
Original file line number Diff line number Diff line change
@@ -0,0 +1,16 @@
##### Azerbaijani #####

>>> Accuracy on average: 85.30%

>> Detection of 1000 single words (average length: 8 chars)
Accuracy: 66.80%
Erroneously classified as Turkish: 13.00%, English: 5.00%, Unknown: 3.00%, German: 2.20%, Esperanto: 0.90%, French: 0.70%, Indonesian: 0.70%, Spanish: 0.70%, Estonian: 0.60%, Finnish: 0.60%, Italian: 0.60%, Polish: 0.50%, Swedish: 0.50%, Bokmal: 0.30%, Croatian: 0.30%, Dutch: 0.30%, Slovene: 0.30%, Albanian: 0.20%, Chinese: 0.20%, Czech: 0.20%, Danish: 0.20%, Hungarian: 0.20%, Malay: 0.20%, Persian: 0.20%, Portuguese: 0.20%, Somali: 0.20%, Afrikaans: 0.10%, Armenian: 0.10%, Basque: 0.10%, Bosnian: 0.10%, Catalan: 0.10%, Greek: 0.10%, Hindi: 0.10%, Mongolian: 0.10%, Romanian: 0.10%, Serbian: 0.10%, Tamil: 0.10%, Urdu: 0.10%

>> Detection of 1000 word pairs (average length: 16 chars)
Accuracy: 89.50%
Erroneously classified as Turkish: 6.60%, English: 0.90%, German: 0.70%, Unknown: 0.40%, Esperanto: 0.30%, Finnish: 0.30%, Spanish: 0.20%, Croatian: 0.10%, Danish: 0.10%, Estonian: 0.10%, French: 0.10%, Indonesian: 0.10%, Italian: 0.10%, Malay: 0.10%, Polish: 0.10%, Serbian: 0.10%, Somali: 0.10%, Swedish: 0.10%

>> Detection of 1000 sentences (average length: 107 chars)
Accuracy: 99.60%
Erroneously classified as Turkish: 0.40%

16 changes: 16 additions & 0 deletions accuracy-reports/fastspell-aggr/Basque.txt
Original file line number Diff line number Diff line change
@@ -0,0 +1,16 @@
##### Basque #####

>>> Accuracy on average: 71.07%

>> Detection of 1000 single words (average length: 9 chars)
Accuracy: 43.60%
Erroneously classified as English: 10.70%, Italian: 5.90%, Spanish: 5.30%, Dutch: 4.50%, Indonesian: 4.20%, Unknown: 4.20%, German: 3.70%, French: 2.60%, Polish: 2.00%, Catalan: 1.90%, Portuguese: 1.90%, Esperanto: 1.80%, Swedish: 1.10%, Romanian: 0.80%, Finnish: 0.70%, Malay: 0.70%, Turkish: 0.60%, Hungarian: 0.50%, Serbian: 0.50%, Croatian: 0.40%, Lithuanian: 0.30%, Bokmal: 0.20%, Czech: 0.20%, Estonian: 0.20%, Macedonian: 0.20%, Swahili: 0.20%, Afrikaans: 0.10%, Albanian: 0.10%, Icelandic: 0.10%, Japanese: 0.10%, Latin: 0.10%, Russian: 0.10%, Slovene: 0.10%, Tagalog: 0.10%, Ukrainian: 0.10%, Urdu: 0.10%, Vietnamese: 0.10%

>> Detection of 1000 word pairs (average length: 17 chars)
Accuracy: 70.10%
Erroneously classified as English: 4.60%, Dutch: 4.20%, Italian: 3.30%, Spanish: 2.90%, German: 2.70%, Unknown: 2.40%, Indonesian: 2.00%, French: 1.30%, Portuguese: 0.90%, Polish: 0.80%, Esperanto: 0.70%, Hungarian: 0.70%, Catalan: 0.60%, Finnish: 0.50%, Bokmal: 0.40%, Swahili: 0.40%, Malay: 0.20%, Swedish: 0.20%, Albanian: 0.10%, Arabic: 0.10%, Bosnian: 0.10%, Croatian: 0.10%, Irish: 0.10%, Latvian: 0.10%, Lithuanian: 0.10%, Serbian: 0.10%, Slovak: 0.10%, Slovene: 0.10%, Turkish: 0.10%

>> Detection of 1000 sentences (average length: 102 chars)
Accuracy: 99.50%
Erroneously classified as Dutch: 0.10%, Italian: 0.10%, Spanish: 0.10%, Swahili: 0.10%, Unknown: 0.10%

16 changes: 16 additions & 0 deletions accuracy-reports/fastspell-aggr/Belarusian.txt
Original file line number Diff line number Diff line change
@@ -0,0 +1,16 @@
##### Belarusian #####

>>> Accuracy on average: 94.97%

>> Detection of 1000 single words (average length: 8 chars)
Accuracy: 87.00%
Erroneously classified as Bulgarian: 2.20%, Russian: 1.80%, Kazakh: 1.50%, Bokmal: 1.40%, Ukrainian: 1.40%, Unknown: 1.30%, Macedonian: 1.20%, Serbian: 1.10%, English: 0.30%, Azerbaijani: 0.10%, Bosnian: 0.10%, Estonian: 0.10%, Italian: 0.10%, Korean: 0.10%, Mongolian: 0.10%, Polish: 0.10%, Turkish: 0.10%

>> Detection of 1000 word pairs (average length: 17 chars)
Accuracy: 97.90%
Erroneously classified as Unknown: 0.50%, Kazakh: 0.40%, Bulgarian: 0.30%, Macedonian: 0.20%, Russian: 0.20%, Turkish: 0.20%, Mongolian: 0.10%, Polish: 0.10%, Ukrainian: 0.10%

>> Detection of 1000 sentences (average length: 105 chars)
Accuracy: 100.00%
Erroneously classified as

16 changes: 16 additions & 0 deletions accuracy-reports/fastspell-aggr/Bengali.txt
Original file line number Diff line number Diff line change
@@ -0,0 +1,16 @@
##### Bengali #####

>>> Accuracy on average: 97.83%

>> Detection of 1000 single words (average length: 7 chars)
Accuracy: 94.10%
Erroneously classified as Unknown: 5.90%

>> Detection of 1000 word pairs (average length: 15 chars)
Accuracy: 99.40%
Erroneously classified as Unknown: 0.60%

>> Detection of 1000 sentences (average length: 87 chars)
Accuracy: 100.00%
Erroneously classified as

16 changes: 16 additions & 0 deletions accuracy-reports/fastspell-aggr/Bokmal.txt
Original file line number Diff line number Diff line change
@@ -0,0 +1,16 @@
##### Bokmal #####

>>> Accuracy on average: 74.63%

>> Detection of 1000 single words (average length: 9 chars)
Accuracy: 55.40%
Erroneously classified as English: 15.10%, German: 6.80%, Italian: 3.40%, Dutch: 2.60%, French: 2.00%, Nynorsk: 1.70%, Spanish: 1.70%, Portuguese: 1.30%, Danish: 1.20%, Unknown: 1.00%, Turkish: 0.80%, Hungarian: 0.70%, Indonesian: 0.60%, Polish: 0.60%, Swedish: 0.60%, Vietnamese: 0.60%, Estonian: 0.50%, Finnish: 0.50%, Croatian: 0.40%, Czech: 0.40%, Afrikaans: 0.30%, Romanian: 0.30%, Slovene: 0.30%, Serbian: 0.20%, Azerbaijani: 0.10%, Bulgarian: 0.10%, Catalan: 0.10%, Icelandic: 0.10%, Irish: 0.10%, Japanese: 0.10%, Latin: 0.10%, Malay: 0.10%, Russian: 0.10%, Ukrainian: 0.10%

>> Detection of 1000 word pairs (average length: 17 chars)
Accuracy: 77.30%
Erroneously classified as English: 6.30%, German: 3.20%, Nynorsk: 3.20%, Italian: 1.90%, Danish: 1.50%, Dutch: 1.40%, Polish: 0.80%, French: 0.70%, Turkish: 0.60%, Finnish: 0.50%, Portuguese: 0.40%, Indonesian: 0.30%, Swedish: 0.30%, Croatian: 0.20%, Czech: 0.20%, Malay: 0.20%, Serbian: 0.20%, Slovene: 0.20%, Spanish: 0.20%, Afrikaans: 0.10%, Latin: 0.10%, Romanian: 0.10%, Unknown: 0.10%

>> Detection of 1000 sentences (average length: 98 chars)
Accuracy: 91.20%
Erroneously classified as Nynorsk: 6.90%, Danish: 0.80%, English: 0.50%, French: 0.20%, Esperanto: 0.10%, Hungarian: 0.10%, Italian: 0.10%, Swedish: 0.10%

16 changes: 16 additions & 0 deletions accuracy-reports/fastspell-aggr/Bosnian.txt
Original file line number Diff line number Diff line change
@@ -0,0 +1,16 @@
##### Bosnian #####

>>> Accuracy on average: 64.70%

>> Detection of 1000 single words (average length: 8 chars)
Accuracy: 54.10%
Erroneously classified as English: 10.50%, Italian: 4.20%, Serbian: 3.60%, Croatian: 3.20%, Czech: 2.30%, Polish: 2.30%, French: 2.00%, Esperanto: 1.80%, Unknown: 1.60%, German: 1.40%, Spanish: 1.40%, Portuguese: 1.20%, Finnish: 1.00%, Hungarian: 1.00%, Slovene: 1.00%, Swedish: 0.80%, Albanian: 0.60%, Slovak: 0.60%, Basque: 0.50%, Bokmal: 0.50%, Dutch: 0.50%, Indonesian: 0.50%, Lithuanian: 0.50%, Romanian: 0.50%, Turkish: 0.50%, Danish: 0.30%, Catalan: 0.20%, Estonian: 0.20%, Latvian: 0.20%, Macedonian: 0.20%, Malay: 0.20%, Icelandic: 0.10%, Latin: 0.10%, Persian: 0.10%, Tagalog: 0.10%, Ukrainian: 0.10%, Vietnamese: 0.10%

>> Detection of 1000 word pairs (average length: 16 chars)
Accuracy: 75.80%
Erroneously classified as Croatian: 6.20%, English: 3.30%, Serbian: 3.30%, Italian: 2.00%, Polish: 1.60%, Esperanto: 0.90%, Slovene: 0.80%, German: 0.60%, Portuguese: 0.60%, Czech: 0.50%, Swedish: 0.50%, French: 0.40%, Slovak: 0.40%, Hungarian: 0.30%, Indonesian: 0.30%, Spanish: 0.30%, Unknown: 0.30%, Albanian: 0.20%, Bokmal: 0.20%, Finnish: 0.20%, Lithuanian: 0.20%, Romanian: 0.20%, Turkish: 0.20%, Catalan: 0.10%, Dutch: 0.10%, Icelandic: 0.10%, Macedonian: 0.10%, Malay: 0.10%, Urdu: 0.10%, Vietnamese: 0.10%

>> Detection of 1000 sentences (average length: 105 chars)
Accuracy: 64.20%
Erroneously classified as Croatian: 24.00%, Serbian: 11.10%, Polish: 0.20%, English: 0.10%, Esperanto: 0.10%, Estonian: 0.10%, Indonesian: 0.10%, Malay: 0.10%

16 changes: 16 additions & 0 deletions accuracy-reports/fastspell-aggr/Bulgarian.txt
Original file line number Diff line number Diff line change
@@ -0,0 +1,16 @@
##### Bulgarian #####

>>> Accuracy on average: 92.33%

>> Detection of 1000 single words (average length: 8 chars)
Accuracy: 83.10%
Erroneously classified as Russian: 4.70%, Ukrainian: 3.60%, Serbian: 3.40%, Unknown: 2.30%, Macedonian: 1.70%, Belarusian: 0.30%, Vietnamese: 0.30%, Kazakh: 0.20%, Mongolian: 0.20%, Croatian: 0.10%, German: 0.10%

>> Detection of 1000 word pairs (average length: 17 chars)
Accuracy: 94.70%
Erroneously classified as Macedonian: 1.70%, Russian: 1.10%, Serbian: 1.10%, Ukrainian: 1.10%, Unknown: 0.20%, Belarusian: 0.10%

>> Detection of 1000 sentences (average length: 89 chars)
Accuracy: 99.20%
Erroneously classified as Russian: 0.70%, Macedonian: 0.10%

16 changes: 16 additions & 0 deletions accuracy-reports/fastspell-aggr/Catalan.txt
Original file line number Diff line number Diff line change
@@ -0,0 +1,16 @@
##### Catalan #####

>>> Accuracy on average: 66.23%

>> Detection of 1000 single words (average length: 8 chars)
Accuracy: 43.50%
Erroneously classified as English: 19.30%, Italian: 6.90%, French: 6.00%, Portuguese: 6.00%, Spanish: 4.30%, Unknown: 2.90%, Swedish: 1.90%, German: 1.70%, Dutch: 1.20%, Esperanto: 0.70%, Indonesian: 0.60%, Bokmal: 0.50%, Polish: 0.50%, Basque: 0.40%, Czech: 0.40%, Malay: 0.40%, Romanian: 0.40%, Turkish: 0.40%, Latin: 0.30%, Croatian: 0.20%, Hungarian: 0.20%, Serbian: 0.20%, Albanian: 0.10%, Armenian: 0.10%, Bengali: 0.10%, Finnish: 0.10%, Lithuanian: 0.10%, Slovak: 0.10%, Tagalog: 0.10%, Tamil: 0.10%, Ukrainian: 0.10%, Urdu: 0.10%, Vietnamese: 0.10%

>> Detection of 1000 word pairs (average length: 16 chars)
Accuracy: 67.30%
Erroneously classified as English: 9.30%, Portuguese: 7.00%, Italian: 4.10%, French: 4.00%, Spanish: 3.60%, Unknown: 1.70%, German: 0.50%, Swedish: 0.50%, Dutch: 0.30%, Indonesian: 0.30%, Polish: 0.30%, Vietnamese: 0.20%, Bokmal: 0.10%, Czech: 0.10%, Esperanto: 0.10%, Estonian: 0.10%, Kazakh: 0.10%, Malay: 0.10%, Russian: 0.10%, Serbian: 0.10%, Slovak: 0.10%

>> Detection of 1000 sentences (average length: 103 chars)
Accuracy: 87.90%
Erroneously classified as Spanish: 5.60%, English: 2.50%, French: 1.40%, Italian: 0.80%, Portuguese: 0.80%, Unknown: 0.50%, Basque: 0.10%, Dutch: 0.10%, German: 0.10%, Japanese: 0.10%, Romanian: 0.10%

16 changes: 16 additions & 0 deletions accuracy-reports/fastspell-aggr/Chinese.txt
Original file line number Diff line number Diff line change
@@ -0,0 +1,16 @@
##### Chinese #####

>>> Accuracy on average: 71.18%

>> Detection of 1000 single words (average length: 1 chars)
Accuracy: 46.30%
Erroneously classified as Japanese: 20.90%, English: 9.10%, Italian: 2.60%, German: 2.50%, French: 2.20%, Korean: 2.00%, Unknown: 1.90%, Russian: 1.70%, Spanish: 1.10%, Swedish: 0.80%, Hungarian: 0.70%, Catalan: 0.60%, Esperanto: 0.60%, Portuguese: 0.60%, Bulgarian: 0.50%, Dutch: 0.50%, Greek: 0.50%, Polish: 0.50%, Turkish: 0.50%, Czech: 0.40%, Finnish: 0.40%, Persian: 0.40%, Serbian: 0.30%, Tamil: 0.30%, Azerbaijani: 0.20%, Danish: 0.20%, Hebrew: 0.20%, Lithuanian: 0.20%, Romanian: 0.20%, Slovak: 0.20%, Ukrainian: 0.20%, Armenian: 0.10%, Bokmal: 0.10%, Hindi: 0.10%, Indonesian: 0.10%, Macedonian: 0.10%, Malay: 0.10%, Vietnamese: 0.10%

>> Detection of 1000 word pairs (average length: 2 chars)
Accuracy: 67.50%
Erroneously classified as Japanese: 23.80%, English: 2.30%, Unknown: 1.30%, Korean: 0.70%, Russian: 0.50%, French: 0.40%, German: 0.40%, Italian: 0.40%, Portuguese: 0.40%, Greek: 0.30%, Turkish: 0.30%, Vietnamese: 0.30%, Armenian: 0.20%, Catalan: 0.20%, Esperanto: 0.20%, Finnish: 0.10%, Hungarian: 0.10%, Lithuanian: 0.10%, Macedonian: 0.10%, Persian: 0.10%, Serbian: 0.10%, Slovene: 0.10%, Spanish: 0.10%

>> Detection of 729 sentences (average length: 48 chars)
Accuracy: 99.73%
Erroneously classified as Japanese: 0.27%

16 changes: 16 additions & 0 deletions accuracy-reports/fastspell-aggr/Croatian.txt
Original file line number Diff line number Diff line change
@@ -0,0 +1,16 @@
##### Croatian #####

>>> Accuracy on average: 81.30%

>> Detection of 1000 single words (average length: 8 chars)
Accuracy: 63.70%
Erroneously classified as English: 8.40%, Italian: 3.20%, Czech: 2.50%, Unknown: 2.50%, Serbian: 2.40%, Polish: 1.80%, French: 1.40%, Swedish: 1.40%, Turkish: 1.40%, Spanish: 1.20%, Esperanto: 1.10%, German: 1.00%, Portuguese: 1.00%, Slovak: 0.90%, Lithuanian: 0.70%, Indonesian: 0.60%, Bokmal: 0.50%, Hungarian: 0.50%, Dutch: 0.40%, Finnish: 0.40%, Latvian: 0.40%, Slovene: 0.40%, Bosnian: 0.30%, Estonian: 0.30%, Romanian: 0.30%, Basque: 0.20%, Korean: 0.20%, Malay: 0.20%, Armenian: 0.10%, Catalan: 0.10%, Chinese: 0.10%, Icelandic: 0.10%, Russian: 0.10%, Tagalog: 0.10%, Vietnamese: 0.10%

>> Detection of 1000 word pairs (average length: 17 chars)
Accuracy: 86.90%
Erroneously classified as English: 2.40%, Italian: 1.50%, Polish: 1.40%, Serbian: 1.00%, French: 0.90%, Czech: 0.80%, Slovak: 0.80%, German: 0.60%, Bokmal: 0.50%, Hungarian: 0.40%, Esperanto: 0.30%, Lithuanian: 0.30%, Slovene: 0.30%, Turkish: 0.30%, Albanian: 0.20%, Malay: 0.20%, Romanian: 0.20%, Swedish: 0.20%, Tagalog: 0.20%, Unknown: 0.20%, Basque: 0.10%, Bosnian: 0.10%, Indonesian: 0.10%, Portuguese: 0.10%

>> Detection of 1000 sentences (average length: 127 chars)
Accuracy: 93.30%
Erroneously classified as Bosnian: 3.30%, Serbian: 2.70%, Polish: 0.40%, English: 0.20%, Slovene: 0.10%

16 changes: 16 additions & 0 deletions accuracy-reports/fastspell-aggr/Czech.txt
Original file line number Diff line number Diff line change
@@ -0,0 +1,16 @@
##### Czech #####

>>> Accuracy on average: 79.83%

>> Detection of 1000 single words (average length: 8 chars)
Accuracy: 64.40%
Erroneously classified as English: 6.00%, Polish: 3.00%, Slovene: 3.00%, Slovak: 2.60%, Italian: 2.30%, Serbian: 1.90%, Hungarian: 1.70%, German: 1.60%, Finnish: 1.30%, Croatian: 1.20%, Spanish: 1.20%, French: 1.10%, Esperanto: 1.00%, Swedish: 0.90%, Unknown: 0.70%, Bosnian: 0.60%, Portuguese: 0.60%, Catalan: 0.50%, Danish: 0.50%, Indonesian: 0.50%, Romanian: 0.50%, Dutch: 0.40%, Estonian: 0.40%, Turkish: 0.40%, Azerbaijani: 0.30%, Bokmal: 0.30%, Latin: 0.20%, Persian: 0.20%, Afrikaans: 0.10%, Icelandic: 0.10%, Irish: 0.10%, Japanese: 0.10%, Latvian: 0.10%, Russian: 0.10%, Welsh: 0.10%

>> Detection of 1000 word pairs (average length: 16 chars)
Accuracy: 82.90%
Erroneously classified as Polish: 3.00%, Slovene: 2.10%, English: 1.60%, Serbian: 1.60%, Slovak: 1.40%, Italian: 0.90%, Croatian: 0.80%, Spanish: 0.70%, Hungarian: 0.50%, Indonesian: 0.50%, Danish: 0.40%, French: 0.40%, German: 0.40%, Portuguese: 0.40%, Esperanto: 0.30%, Estonian: 0.30%, Romanian: 0.30%, Unknown: 0.30%, Bosnian: 0.20%, Dutch: 0.20%, Turkish: 0.20%, Basque: 0.10%, Catalan: 0.10%, Finnish: 0.10%, Latin: 0.10%, Latvian: 0.10%, Swedish: 0.10%

>> Detection of 1000 sentences (average length: 93 chars)
Accuracy: 92.20%
Erroneously classified as Polish: 1.50%, Slovak: 1.50%, English: 1.40%, Slovene: 1.10%, German: 0.40%, Portuguese: 0.30%, Serbian: 0.30%, Italian: 0.20%, Latin: 0.20%, Bosnian: 0.10%, Esperanto: 0.10%, Finnish: 0.10%, Hungarian: 0.10%, Persian: 0.10%, Spanish: 0.10%, Swedish: 0.10%, Turkish: 0.10%, Unknown: 0.10%

16 changes: 16 additions & 0 deletions accuracy-reports/fastspell-aggr/Danish.txt
Original file line number Diff line number Diff line change
@@ -0,0 +1,16 @@
##### Danish #####

>>> Accuracy on average: 78.03%

>> Detection of 1000 single words (average length: 8 chars)
Accuracy: 57.70%
Erroneously classified as English: 15.40%, German: 7.70%, Dutch: 2.80%, French: 2.60%, Italian: 1.90%, Spanish: 1.50%, Polish: 0.80%, Unknown: 0.80%, Czech: 0.70%, Afrikaans: 0.60%, Bokmal: 0.60%, Finnish: 0.60%, Portuguese: 0.60%, Hungarian: 0.50%, Nynorsk: 0.50%, Turkish: 0.50%, Esperanto: 0.40%, Estonian: 0.40%, Indonesian: 0.40%, Swedish: 0.40%, Basque: 0.30%, Slovak: 0.30%, Vietnamese: 0.30%, Catalan: 0.20%, Latin: 0.20%, Romanian: 0.20%, Slovene: 0.20%, Albanian: 0.10%, Azerbaijani: 0.10%, Bosnian: 0.10%, Chinese: 0.10%, Greek: 0.10%, Icelandic: 0.10%, Japanese: 0.10%, Malay: 0.10%, Russian: 0.10%

>> Detection of 1000 word pairs (average length: 16 chars)
Accuracy: 77.50%
Erroneously classified as English: 8.00%, German: 5.00%, Dutch: 1.70%, French: 1.50%, Finnish: 1.00%, Italian: 0.80%, Bokmal: 0.60%, Nynorsk: 0.60%, Estonian: 0.50%, Slovene: 0.40%, Spanish: 0.40%, Turkish: 0.40%, Czech: 0.30%, Afrikaans: 0.20%, Esperanto: 0.20%, Indonesian: 0.20%, Polish: 0.20%, Unknown: 0.20%, Basque: 0.10%, Bosnian: 0.10%, Hungarian: 0.10%

>> Detection of 1000 sentences (average length: 112 chars)
Accuracy: 98.90%
Erroneously classified as Nynorsk: 0.30%, Bokmal: 0.20%, Dutch: 0.20%, English: 0.10%, French: 0.10%, Hungarian: 0.10%, Turkish: 0.10%

Loading

0 comments on commit 3f1359c

Please sign in to comment.