Skip to main content

Table 8 Measures (M) of Precision (Pr), Recall (Rc) and \(\hbox {F}_1\) scores for Task D and Task E

From: Paraphrase type identification for plagiarism detection using contexts and word embeddings

Embedding /Alignment

M

Task D

Task E

ConceptNet Numberbatch

FastText

GloVe

ConceptNet Numberbatch

FastText

GloVe

SW Algo.

Pr

0.88186

0.78737

0.81406

0.76623

0.70969

0.73386

Rc

0.89947

0.83510

0.83050

0.76770

0.76360

0.68662

\(\hbox {F}_1\)

0.89058

0.81053

0.82220

0.76696

0.73566

0.70945

Meteor

Pr

0.79964

0.72269

0.74560

0.71632

0.64403

0.66710

Rc

0.83633

0.80790

0.79422

0.71683

0.69614

0.65296

\(\hbox {F}_1\)

0.81757

0.76293

0.76914

0.71657

0.66907

0.65995

Sultan

Pr

0.79733

0.71772

0.74089

0.72821

0.64249

0.67028

Rc

0.78477

0.76078

0.74266

0.68411

0.67920

0.62313

\(\hbox {F}_1\)

0.79120

0.73862

0.74177

0.70563

0.66034

0.64585

  1. Bold values refer to the highest values of \(\hbox {F}_1\) scores of each task