Figure 8
From: Evaluating gold standard corpora against gene/protein tagging solutions and lexical resources
The diagram gives an overview on the most frequent FN results. The FN results have been categorized according to the morphology and the semantics of the missed terms. For all GSCs and the different tagging solutions the counts for the 15 most frequent FN errors are shown. Since PennBio and FsuPrge are the largest GSCs, they induce also a larger number of FN results. FN results of GE terms are mainly encountered in Jnlpba. FsuPrge generates a larger portion of PGT FN results, whereas PennBio mainly leads to a-PG FN results.