Extracting Medical Knowledge from Crowdsourced Question Answering Website (Preprint-2017)


The main aim of this paper is to extract the high quality knowledge from the noisy question answer pairs in online medical Q/A websites and at the same time, estimate expertise for the doctors who give answers on these Q&A websites. For this propose we developed a new system called opinion Target Finding (OPF).In this System the exact answer for a question will be found out without any supervision and the answer will be a trust Worthy one.

Proposed system:

In our project we develop a approach based on the UN-supervised model, which regards identifying correct trustworthy answer. To fulfill this aim, both Questions and answers must be under consideration. First, however, it is necessary to find the question and the question topics in which the patient and the doctor is communicating. For an Example

Question: I am having severe throat pain and body pain
Answer1: May be the Symptoms of fever
Answer2: May be the symptoms of throat infection
Answer3: May be symptoms of cold

The types of modules identified from the proposed system
• Authentication & Posting Questions
• Stemming
• Trustworthy Calculation
• Report

Separate login will be provided for Patient, Admin and Doctors. They will be using their mailed and password.
When a Patient is verified he\she will be able to post their symptoms to the websites.
This will be view to all doctors as well as patients .The Answers will be posted by any doctors Text analysis involves potter stemmer algorithm which removes the unwanted words from the sentence and word analyzer are used to find Medical words from the database.
The confidence will be calculated using the trustworthy calculation