A Method for Accuracy Improvement with Ontology in Automatic Summarization of Academic Articles

Abstract

This study aims to provide a new method to improve accuracy of a summary by employing ontology in automatic summarization of academic articles focused on segment structure. Existing automatic summarization methods cannot recognize the meaning or interrelationships of the meaning of words. Ontology can recognize them and it is expected to enable more precise choice of important sentences. Specifically, this study aims to increase accuracy of choosing important sentences that are suitable for each segment by assigning ontological characteristics to each segment. Also, this study aims to improve overall quality of the summary by adjusting sentences picked up from each segment, considering interrelationships of the meaning of words.