Dear community!
I am currently searching for a text dataset and maybe you can point me in the right direction.
I am looking for a dataset that possibly covers a medical problem (not necessarily), addresses (multilabel) document classification and contains some numeric values in the text. Ideally, the numeric values (e.g. measurements) have an influence on the classification labels (e.g. label is 'fever' when temperature in the text is above X). Of course, large data and many labels would be great, but I am thankful for interesting suggestions.
Thank you!
Roland
Dear Roland Roller ,
Emergency departments (EDs) in hospitals are usually overcrowded by patients with various severity of health problems1. Prompt diagnosis and therapeutical decisions in such streaming environment might not be always based on the best available expert opinions. For example, patients with potential limb fractures may be treated by only examining the radiology images and prior to the availability of reports written by the radiologists. The manual reconciliation of the initial diagnosis with the formal radiology report usually occurs after the patient has been discharged from the ED2. As a result, timely follow-up of patients with abnormalities in their reports is a critical task.
Article Clinical Document Classification Using Labeled and Unlabeled...
Regards,
Shafagat
Thank you, Shafagat Mahmudova! And yes, I was thinking of this kind of data. The question is, is this data be publicly available? It doesn't seems so, reading quickly over the paper you suggested.
All the best,
Roland
Check UCI or Kaggle machine Learning databases, numerous datasets are there. You may find what suits your problem.
- Badges
- Science topic
- Similar topics
- Organizational Psychology
- Training