Table 1 Sensitivity of gene name extraction of PSE. Five datasets were used for the evaluation. The first two datasets represented as Set 1A, 1B contain forty abstracts retrieved with randomly generated PubMed IDs, respectively. The next two datasets labeled as Set 2A, 2B contain forty abstracts which were randomly selected from the 4,548 abstracts retrieved with "gene AND disease AND activation" as keywords, respectively. The last dataset represented as GENIA is the result from using the GENIA corpus containing 2,000 abstracts. The results using Set 1A,1B and Set 2A, 2B and GENIA are represented as Ev1, Ev2 and Ev3 in this manuscript, respectively. TP, FP and FN represent the true positive, the false positive and the false negative, respectively.
Dataset | TP | FP | FN | Precision | Recall | F-measure |
|---|---|---|---|---|---|---|
Set 1A | 50 | 15 | 61 | 76.9% | 45.0% | 56.8% |
Set 1B | 40 | 11 | 21 | 78.4% | 65.6% | 71.4% |
Set 2A | 287 | 23 | 157 | 92.6% | 64.6% | 76.1% |
Set 2B | 291 | 49 | 210 | 85.6% | 58.1% | 69.2% |
GENIA | 12,842 | 7,752 | 6,912 | 62.4% | 65.0% | 63.7% |