PropertyValue
?:about
?:abstract
  • In this study, we address the problem of paragraph segmentation from the perspective of understanding the content of a novel. Estimating the paragraph of a text can be considered a binary classification problem regarding whether two given sentences belong to the same paragraph. When the number of paragraphs is small relative to the number of sentences, it is necessary to consider the imbalance in the number of data. We applied the bidirectional encoder representations from transformer (BERT), which has shown high accuracy in various natural language processing tasks, to paragraph segmentation. We improved the performance of the model using the focal loss as the loss function of the classifier. As a result, the effectiveness of the proposed model was confirmed on multiple datasets with different ratios of data in each class. (xsd:string)
?:author
?:contributor
?:hasFulltext
  • true (xsd:boolean)
is ?:hasPart of
?:inLanguage
  • Englisch (EN) (xsd:string)
?:libraryLocation
?:name
  • Improving BERT with Focal Loss for Paragraph Segmentation of Novels (xsd:string)
?:provider
?:publicationType
  • Buch (de)
  • Elektronische Ressource (xsd:string)
  • book (en)
?:sourceInfo
  • GESIS-BIB (xsd:string)
rdf:type