From a random sample of 500 companion animal cases of diarrhea, the accuracy of the text miner for classifying the cases as positive or negative for ‘<i>had diagnostic testing</i>’ when compared to a manual review of the medical records serving as the external standard.

<p>From a random sample of 500 companion animal cases of diarrhea, the accuracy of the text miner for classifying the cases as positive or negative for ‘<i>had diagnostic testing</i>’ when compared to a manual review of the medical records serving as the external standard.