Treebank Statistics: UD_Korean-Kaist: POS Tags: SCONJ
There are 6729 SCONJ lemmas (7%), 6706 SCONJ types (7%) and 17587 SCONJ tokens (5%).
Out of 17 observed tags, the rank of SCONJ is: 5 in number of lemmas, 5 in number of types and 6 in number of tokens.
The 10 most frequent SCONJ lemmas: 아니+라, 의하+어, 위하+어, 따르+아, 같+이, 통하+어, 크+게, 보+면, 대하+어, 없+이
The 10 most frequent SCONJ types: 아니라, 따라, 의해, 같이, 위해, 크게, 보면, 통해, 없이, 대해
The 10 most frequent ambiguous lemmas: 아니+라 (SCONJ 445, CCONJ 6, ADJ 1), 따르+아 (SCONJ 323, VERB 2, CCONJ 1), 통하+어 (SCONJ 262, VERB 2), 대하+어 (SCONJ 221, VERB 2), 이렇+게 (SCONJ 134, VERB 23), 어떻+게 (SCONJ 122, VERB 18), 들+어 (SCONJ 104, VERB 38), 있+어 (SCONJ 92, ADJ 4), 하+어 (SCONJ 89, VERB 29), 그렇+게 (SCONJ 87, VERB 21)
The 10 most frequent ambiguous types: 아니라 (SCONJ 445, CCONJ 6, ADJ 2), 따라 (SCONJ 324, VERB 3, CCONJ 1), 같이 (SCONJ 273, ADV 26), 크게 (SCONJ 250, ADV 3), 대해 (SCONJ 155, VERB 3), 있어서 (SCONJ 137, AUX 29, NOUN 1), 이렇게 (SCONJ 136, ADV 54, VERB 23), 어떻게 (SCONJ 122, ADV 36, VERB 18), 해도 (SCONJ 107, AUX 11, NOUN 1), 들어 (SCONJ 105, VERB 40)
- 아니라
- 따라
- 같이
- 크게
- 대해
- 있어서
- 이렇게
- 어떻게
- 해도
- 들어
Morphology
The form / lemma ratio of SCONJ is 0.996582 (the average of all parts of speech is 0.998034).
The 1st highest number of forms (3) was observed with the lemma “관련+하+어”: 견련하여, 관련하여, 관련해.
The 2nd highest number of forms (3) was observed with the lemma “나누+어”: 나누어, 나눠, 나뉘어.
The 3rd highest number of forms (3) was observed with the lemma “내+어”: 내, 내어, 소리내어.
SCONJ does not occur with any features.
Relations
SCONJ nodes are attached to their parents using 18 different relations: ccomp (9885; 56% instances), xcomp (3835; 22% instances), conj (1488; 8% instances), root (854; 5% instances), mark (662; 4% instances), fixed (322; 2% instances), amod (167; 1% instances), advcl (121; 1% instances), cc (75; 0% instances), nmod (59; 0% instances), obj (41; 0% instances), obl (23; 0% instances), dislocated (19; 0% instances), acl (16; 0% instances), nsubj (9; 0% instances), dep (7; 0% instances), case (3; 0% instances), appos (1; 0% instances)
Parents of SCONJ nodes belong to 12 different parts of speech: VERB (12295; 70% instances), CCONJ (1698; 10% instances), SCONJ (1453; 8% instances), (854; 5% instances), NOUN (556; 3% instances), ADJ (366; 2% instances), ADV (344; 2% instances), PROPN (12; 0% instances), PART (5; 0% instances), NUM (2; 0% instances), INTJ (1; 0% instances), X (1; 0% instances)
4233 (24%) SCONJ nodes are leaves.
7787 (44%) SCONJ nodes have one child.
3581 (20%) SCONJ nodes have two children.
1986 (11%) SCONJ nodes have three or more children.
The highest child degree of a SCONJ node is 6.
Children of SCONJ nodes are attached using 22 different relations: obj (3889; 18% instances), obl (3365; 16% instances), advcl (2551; 12% instances), nsubj (2360; 11% instances), advmod (1914; 9% instances), conj (1683; 8% instances), dislocated (1416; 7% instances), ccomp (1349; 6% instances), punct (1060; 5% instances), nmod (432; 2% instances), xcomp (409; 2% instances), dep (385; 2% instances), csubj (246; 1% instances), amod (188; 1% instances), compound (169; 1% instances), iobj (122; 1% instances), nummod (48; 0% instances), cc (28; 0% instances), det (17; 0% instances), acl (9; 0% instances), discourse (8; 0% instances), mark (2; 0% instances)
Children of SCONJ nodes belong to 15 different parts of speech: NOUN (7911; 37% instances), ADV (7270; 34% instances), SCONJ (1453; 7% instances), VERB (1380; 6% instances), PUNCT (1060; 5% instances), CCONJ (857; 4% instances), ADJ (641; 3% instances), PRON (467; 2% instances), PROPN (442; 2% instances), NUM (107; 0% instances), SYM (27; 0% instances), DET (17; 0% instances), INTJ (8; 0% instances), PART (8; 0% instances), X (2; 0% instances)