An improved chatbot for predicting disease and medicines using natural language processing with fuzzy logic

Abstract

The aim is to create an artificial conversation entity(chatbot) using python to predict disease and medicine for healthcare treatments. Two algorithms fuzzy support vector machine algorithms are compared with Decision tree algorithm sample size taken 28. G power of 81% and sample size is calculated using the G power tool. Performances of the score model validated test set accuracy with 95% confidence interval for fuzzy support vector machine algorithm with different sub-samples has 91.60% accuracy comparing with Decision tree which has 87.90% accuracy.Independent Sample T-test a significance difference in accuracy and loss is observed p<0.005.From the results it is concluded that proposed algorithm Fuzzy support vector machine will produce better results than the existing algorithm.