Recommendation Algorithm Based on Knowledge Graph to Propagate User Preference

Abstract

Applications of machine learning in the biological and healthcare fields have improved early disease identification and diagnosis. This has recently improved patient treatment. Studies have revealed that consumers turn to the internet for advice on any potential health-related problems. This method has a drawback in that the search engines offer a lot of information in a disorganised way that makes it challenging to draw conclusions from. The game of disease detection based on disease is intricate. The users feed the symptoms in non-technical or natural terms because they are unfamiliar with biological terms, which makes disease prediction more difficult. The main goal is to create a novel architecture capable of accepting and handling these kinds of user queries by utilising methods like query expansion using a thesaurus, synonym matching, and symptom suggestion that will enable disease prediction with a higher degree of accuracy based on user input and recommend the best treatment possible for that disease. We collected data from the internet and created a dataset that can be applied to further study. In these situations, prediction is accomplished using matching and query search retrieval.