Comparison of Data Augmentation Methods in Pointer-Generator Model Using Various Sentence Ranking Methods

In the existing research, we proposed a data augmentation method using topic model for Pointer-Generator model. In this study, we add to the sentence ranking method in the data augmentation method. Specifically, we add two ranking methods using LexRank and Luhn. LexRank is based on Google's search method and Luhn defines sentence features and ranks sentences. We compare three data augmentation methods. We considered which method is suitable for data augmentation. We confirm that most accurate model is the model using data augmentation method by topic model.