INTERNATIONAL JOURNAL OF SCIENTIFIC DEVELOPMENT AND RESEARCH International Peer Reviewed & Refereed Journals, Open Access Journal ISSN Approved Journal No: 2455-2631 | Impact factor: 8.15 | ESTD Year: 2016
open access , Peer-reviewed, and Refereed Journals, Impact factor 8.15
Admittance to emergency clinic information is usually a difficult, exorbitant and tedious cycle requiring broad cooperation with network overseers. This prompts potential deferrals in obtaining bits of knowledge from information, like conclusion or other clinical results. Medical care managers, clinical professionals, specialists and patients could benefit from a framework that could separate significant data from medical services information in constant. In this paper, we present an inquiry addressing framework that permits wellbeing experts to interface with a enormous scope information base by posing inquiries in normal language. This framework is based upon the BERT and SQLOVA models, which make an interpretation of a client’s solicitation into a SQL question, which is then passed to the information server to recover pertinent information. We additionally propose a profound bilinear comparability model to work on the created SQL inquiries by better matching terms in the client’s question with the information base mapping and substance. This framework was prepared with just 75 genuine inquiries and 455 backinterpreted inquiries, and was assessed north of 75 additional genuine inquiries regarding a genuine wellbeing data set, accomplishing a recovery precision of 78 percent.
Keywords:
Medical Chatbot, Django, NLP, BERT
Cite Article:
"Medical Chatbot Using Question Answering Model", International Journal of Science & Engineering Development Research (www.ijsdr.org), ISSN:2455-2631, Vol.7, Issue 5, page no.473 - 475, May-2022, Available :http://www.ijsdr.org/papers/IJSDR2205091.pdf
Downloads:
000337074
Publication Details:
Published Paper ID: IJSDR2205091
Registration ID:200465
Published In: Volume 7 Issue 5, May-2022
DOI (Digital Object Identifier):
Page No: 473 - 475
Publisher: IJSDR | www.ijsdr.org
ISSN Number: 2455-2631
Facebook Twitter Instagram LinkedIn