Treebank Statistics: UD_Marathi-UFAL: POS Tags: NUM
There are 14 NUM lemmas (2%), 16 NUM types (1%) and 34 NUM tokens (1%).
Out of 16 observed tags, the rank of NUM is: 7 in number of lemmas, 9 in number of types and 13 in number of tokens.
The 10 most frequent NUM lemmas: दोन, हजार, एक, चार, पाचपन्नास, चाळीस, चौघे, दहा, दुसरा, पाच
The 10 most frequent NUM types: दोन, हजार, एक, चार, दुसरा, पाचपन्नास, चाळीस, चौघे, दहा, दुसऱ्या
The 10 most frequent ambiguous lemmas: एक (DET 28, NUM 3), चौघे (PRON 2, NUM 1), दुसरा (ADJ 1, NUM 1)
The 10 most frequent ambiguous types: एक (DET 23, NUM 3), चौघे (PRON 2, NUM 1)
- एक
- चौघे
Morphology
The form / lemma ratio of NUM is 1.142857 (the average of all parts of speech is 1.560265).
The 1st highest number of forms (3) was observed with the lemma “दोन”: दुसरा, दोन, दोन्ही.
The 2nd highest number of forms (1) was observed with the lemma “एक”: एक.
The 3rd highest number of forms (1) was observed with the lemma “चार”: चार.
NUM occurs with 3 features: Case (3; 9% instances), Number (3; 9% instances), Gender (1; 3% instances)
NUM occurs with 3 feature-value pairs: Case=Nom, Gender=Neut, Number=Plur
NUM occurs with 3 feature combinations.
The most frequent feature combination is _ (31 tokens).
Examples: दोन, हजार, एक, चार, दुसरा, पाचपन्नास, चाळीस, दहा, दुसऱ्या, दोन्ही
Relations
NUM nodes are attached to their parents using 4 different relations: nummod (31; 91% instances), amod (1; 3% instances), flat (1; 3% instances), obj (1; 3% instances)
Parents of NUM nodes belong to 5 different parts of speech: NOUN (29; 85% instances), NUM (2; 6% instances), ADJ (1; 3% instances), ADV (1; 3% instances), VERB (1; 3% instances)
29 (85%) NUM nodes are leaves.
5 (15%) NUM nodes have one child.
The highest child degree of a NUM node is 1.
Children of NUM nodes are attached using 4 different relations: amod (2; 40% instances), flat (1; 20% instances), nummod (1; 20% instances), punct (1; 20% instances)
Children of NUM nodes belong to 3 different parts of speech: ADJ (2; 40% instances), NUM (2; 40% instances), PUNCT (1; 20% instances)