Skip to main content

Table 9 F-score performance of the inter- and intra corpus experiments according to the exact matching criterion

From: Investigating heterogeneous protein annotations toward cross-corpora utilization

Testing data Training data
  AIMed GENETAG GENIA
AIMed 77.68 69.34 66.16
GENETAG 68.96 69.65 63.62
GENIA 78.36 77.80 78.76
  1. "Testing data" represents the corpus used for evaluation. "Training data" represents the corpus which was "added" to the training portion of the "Testing data" corpus for training. For example, the score in the cell AIMed-GENETAG represents the performance of training on the training portion of the AIMed and GENETAG corpora, and testing on the testing portion of AIMed.