Book Topic and Emotion Classification by Child Readers in a Library in Taiwan

This study created a visualized folksonomy classification that integrates topic icons and emojis to help children to find books that they like in library settings. The National Library of Public Information in Taiwan commissioned a book-navigation app using the proposed classification. We recruited 35 children to use this app to search for books to read, and they recorded their feelings about these books over a twelve-week period. A statistical analysis of 1,938 system logs was performed under three themes: thematic preferences, epistemic cognition, and social communication. The implementation of thematic and emotion icons in a folksonomy topic structure appeared to successfully aid young Taiwanese readers in finding books that they wanted. Evidence of complex relationships among cultural, psychological, cognitive, and social aspects of book-finding was found, suggesting the benefits of supervised learning in the extraction of gross-fine emotions in the sentiment analysis of multiple topics.