A Case-Based Reasoning for Detection Coronavirus (Covid-19) Using Cosine Similarity

Abstract

Case-based reasoning is a new approach that can be used to diagnose disease in addition to using expert systems or other approaches, which are part of artificial intelligence. Case-based reasoning can diagnose diseases based on visible or perceived clinical symptoms. This study tries to build case-based reasoning for early detection of COVID-19 by looking at the characteristics of clinical symptoms seen in a person using the Cosine Similarity method. Cosine similarity is a method to find level of similarity between two cases. The detection process is carried out by entering a new case containing symptoms into the system, then system will perform a similarity calculation process between the old case and the new case. The results show case-based reasoning for early detection of COVID-19 using the Cosine Similarity method can detect a similarity level of 80%.