Bengali Medical Dataset for Multi-purpose
Description
This dataset has been created for our country, Bangladesh where people will get help from it. This dataset is for the medical specialist classification and Bengali Named Entity Recognition which will play a vital role in multi-purpose.
Files
Steps to reproduce
It is an NLP dataset in the medical sector with around six hundred patient primary statements of pure Bengali language, around eight thousand words. The data were created according to the statements of the medical outdoor receptionists and some doctors’ experiences about what kind of problems the patients come to the doctors and share their health issues. The patients come first and evolve their problems and by their statements, the receptionist helps them to notify about which specialist they should consult. All data were verified by an authorized MBBS doctor. Also, data were collected through maximum patients’ statements suffering from different problems. This dataset is created with the general and common health issues of as usually people face and also the primary symptoms of any major problems like headache, bone pain, itching problem, etc. This dataset is partitioned into two parts and annotated manually. One part is labeled with patient symptoms, body parts, colors, body fluids, blood, times, values, directions, effluents, adverbs, and another part is labeled with specialists like medicine specialist, cardiologist, dentist, gynecologist, etc. This dataset will be helpful especially in the upcoming period as every health complex, health centers, hospitals all over the world are facing a very rush period day by day.
Institutions
- Bangladesh University of Business and Technology