Table 1.
Performance of each tool.
| Types | Tools | Date | Prec. | Recall | F-score | Choice | DataSet | Time* |
|---|---|---|---|---|---|---|---|---|
| Chemical | CheNER | 2013 | 91% | 30% | 50% | O | EvalC Corpus | 21 min |
| ChemSpot | 2012 | 75% | 84% | 79% | X | 210 min | ||
|
Gene/ Protein |
GENIA tagger | 2006 | 66% | 81% | 72% | O | NLPBA2004 (Protein) |
9 min |
| NERsuite | 2012 | 88% | 82% | 85% | X | BioCreative2 | 11 min | |
| ABNER | 2005 | 68% | 78% | 73% | X | |||
| BANNER | 2008 | 87% | 83% | 85% | X | 60 min | ||
|
Alias-i LingPipe (LIPI) |
2010 | 60% | 70% | 64% | X | |||
| Cell type | GENIA tagger | 2006 | 79% | 71% | 74% | O | NLPBA dataset | 9 min |
| ABNER | 2005 | 80% | 66% | 72% | X | |||
| Disease | BANNER | 2008 | 87% | 83% | 85% | O | NCBI Disease Corpus | 58 min |
| Organ |
Alias-i LingPipe (LIPI) |
2010 | O | NCBI's GeneTag Corpus | 16 min | |||
Time* is time for which it takes to tag entities within 5,000 abstracts.