Skip to main content

Table 3 Gene mention tagger performance for six combinations of tagger and model.

From: The structural and content aspects of abstracts versus bodies of full text journal articles are different

   

Abstracts

     

Bodies

  

Tagger and model

P

R

F

TP

FP

FN

P

R

F

TP

FP

FN

Human annotations

   

1180

     

22,432

  

ABNER/BioCreative

0.634

0.393

0.485

464

267

716

.505

.322

.394

7,240

7,081

15,186

ABNER/NLPBA

0.629

0.363

0.460

429

253

751

.464

.298

.363

6,694

7,719

15,730

BANNER/BioCreative

0.678

0.482

0.563

569

270

611

.540

.473

.504

10,621

9,042

11,806

LingPipe/BioCreative

0.591

0.575

0.583

679

468

501

.353

.583

.440

13,094

23,992

9,330

LingPipe/GENIA

0.398

0.309

0.348

365

551

815

.250

.284

.266

6,380

19,101

16,046

LingPipe/JNLPBA

0.464

0.277

0.347

327

377

853

.289

.256

.271

5,741

10,498

16,684

  1. The first row gives the number of gene mentions in each set of texts.
Back to article page

BMC Bioinformatics

ISSN: 1471-2105

Contact us