Some english medical documents are currently also detected as positive (probably because of the shared latin vocabulary). So we either need a more complex model which is able of differentiating between "english medical" and "german medical" - or we need a preclassification stage for language discrimination.
Some english medical documents are currently also detected as positive (probably because of the shared latin vocabulary). So we either need a more complex model which is able of differentiating between "english medical" and "german medical" - or we need a preclassification stage for language discrimination.