Skip to Main Content
Table 4 

Classification results on CLEF-IP 2010 English abstracts, with ranges for a 95% confidence interval. Bold figures indicate the best results obtained with the five classifiers. (P: Precision; R: Recall, F1: F1-score).


P
R
F1
weighted random guessing 6.09% ± 0.14 6.04% ± 0.14 6.06% ± 0.14 
unigrams 76.27% ± 0.26 66.13% ± 0.28 70.84% ± 0.27 
unigrams + bigrams 79.00% ± 0.24 70.19% ± 0.27 74.34% ± 0.26 
unigrams + Stanford triples 78.35% ± 0.25 69.57% ± 0.28 73.70% ± 0.26 
unigrams + AEGIR triples 78.51% ± 0.25 69.18% ± 0.28 73.55% ± 0.26 
all representations 79.51% ± 0.24 71.11% ± 0.27 75.08% ± 0.26 

P
R
F1
weighted random guessing 6.09% ± 0.14 6.04% ± 0.14 6.06% ± 0.14 
unigrams 76.27% ± 0.26 66.13% ± 0.28 70.84% ± 0.27 
unigrams + bigrams 79.00% ± 0.24 70.19% ± 0.27 74.34% ± 0.26 
unigrams + Stanford triples 78.35% ± 0.25 69.57% ± 0.28 73.70% ± 0.26 
unigrams + AEGIR triples 78.51% ± 0.25 69.18% ± 0.28 73.55% ± 0.26 
all representations 79.51% ± 0.24 71.11% ± 0.27 75.08% ± 0.26 
Close Modal

or Create an Account

Close Modal
Close Modal