COVERAGE STATISTICS

The ATB1v3.0 (Arabic Treebank: Part 1 v3.0) Corpus contains 145,386 tokens, 
of which 21,591 are punctuation, numbers, and Latin strings, and 123,795 
are Arabic word tokens. 

Punctuation, Numbers, Latin strings       21,591
Arabic Word Tokens                       123,795
------------------------------------------------
TOTAL                                    145,386


Of the 123,795 Arabic word tokens, 233 were found to be typos in the original 
text. Of the remaining 123,562 Arabic word tokens, 123,192 (99.70%) were 
provided with an accurate morphological analysis and POS tag by the 
Buckwalter Arabic Morphological Analyzer, and 370 (0.30%) were judged to be 
incorrectly analyzed, and were flagged with a Comment describing the nature 
of the inaccuracy.

Accurately parsed Arabic Word Tokens        123,192   99.70%
Commented Arabic Word Tokens                    370    0.30%
------------------------------------------------------------
TOTAL                                       123,562  100.00%


The Annotator Comments show that most of the words that were not provided 
with a correct morphological analysis were words missing in the Analyzer 
lexicon, passive verbs, and words with missing hamza (which could be 
considered a form of typo). 

============================================
ANNOTATOR COMMENTS ON ITEMS WITH NO SOLUTION
============================================
NOT_IN_LEXICON (MISC)             204  55.14
PASSIVE_FORM                       76  20.54
MISSING_HAMZA_PROBLEM              67  18.11
NOUN_SHOULD_BE_ADJ                  9   2.43
GRAMMAR_PROBLEM                     7   1.89
DIALECTAL_FORM                      5   1.35
ADJ_SHOULD_BE_NOUN                  2   0.54
--------------------------------------------
TOTAL                             370 100.00
============================================