Or @rebshoe et al. (2013): "These results show that the tagger performances highly depend on the corpora that they have been trained on and thus they cannot generalize well from the learning corpora to the other corpora." (for gene names) https://t.co/hQO
Old news (see @rebshoe https://t.co/S7IV92EOIr) but good to have verification: "We demonstrate that accuracies of models trained on individual corpora decrease substantially for recognition of the same biomolecular entity classes in independent corpora."
Nice Rebholz-Schuhmann paper: for gene names machine learning beats dictionary matching but it is not a home run. http://t.co/1Ap1OjMDcy
Nice Rebholz-Schuhmann paper: for gene names machine learning beats dictionary matching but it is not a home run. http://t.co/1Ap1OjMDcy