부산대학교 도서관

Free-text is categorised into predetermined groups in text classification, a crucial field of machine learning. Due to its many uses, such as sentiment analysis, topic labelling, language identification, and spam filtering, it has attracted a great deal of study interest. Ontologies and machine learning give the knowledge of data relations or patterns, which greatly increases the text categorization efficiency. The goal of this study is to diagnose myocardial infraction among patients by machine learning classifiers. The code that goes with the created approach is meant to be flexible and work with any text categorization system. In this research, we present an improved architecture that makes use of ontology to improve the efficacy of text categorization and diagnose myocardial infraction. In particular, we provide a novel ontology-driven text categorization technique that leverages the extensive semantic data from the Disease Ontology (DOID). We examine previous research and argue that ontology-based text categorization strategies are better than traditional techniques in a number of criteria, including as F-measure, Accuracy, Precision, and Recall.