Table 3:
Illustration of features on the sentence Our data are generally consistent with that of other studies [4–7], but not with studies where a single dose of paracetamol was administered, which reported no significant impact on any reactions [8,9]. The citation under consideration for this example is [4–7] and the citation context is the current sentence only.
Feature Type | Features |
---|---|
n-gram features | |
n-grams (ngram) | [oc], [tc], administered, which_reported, are_generally_consistent, … |
Lemma n-grams (ngraml) | [oc], [tc], administer, which_report, be_generally_consistent, … |
POS n-grams(posgram) | cc, dt, nn, cc_rb, dt_in, cc_rb_in, … |
POS-augmented lemma n-grams (ngraml_pos) | [oc], [tc], administer_vb, which_wd__report_vb, be_vb__generally_rb__consistent_jj … |
Negation count | 2 (due to no, not) |
Negated unigrams | not_impact, not_with, not_study, not_significant |
Sentiment lexicon features (sent) | |
POS sentiment | true (due to consistent with) |
NEG sentiment | false |
Any sentiment | true |
Structure features (struct) | |
n-grams | [oc], [tc], citingwork, pos, contrast, citingwork_pos, citingwork_pos_[tc], … |
Direction | citingwork_contrast, [tc]_contrast, … |
Additional features | |
Dependency features (dep) | advcl_consistent_administer, neg_consistent_not, nsubj_consistent_datum, … |
Rule-based output (rule) | pos |