Skip to main content
Figure 1 | Journal of Biomedical Semantics

Figure 1

From: Evaluating gold standard corpora against gene/protein tagging solutions and lexical resources

Figure 1

The tagging solutions have been used to annotated the lexical resource to determine how they comply with naming standards. The figure shows the tagging of entries from larger dictionary resources (SwissProt (GP7), GP7) against the smaller lexical resource (SwissProt). When using exact matching, only about 53% and 60% of the terms in the smaller resource can be identified by SwissProt (GP7) and GP7, respectively; whereas 88% and 90%, respectively, have been tagged when using alias matching. The same results are produced when using the small lexical resource as tagger against the larger lexical resources. The first experiment using exact matching demonstrates that the larger lexical resource uses slightly different notation standards which can be ignored when using alias matching. It is also remarkable that the smaller lexical resource seems to be already very effective in the tagging of the genes in the corpora indicating that the core terminology for gene mentions is already included in a comprehensive way.

Back to article page