Fiction Popularity Prediction Based on Emotion Analysis

In addition to bringing us knowledge, books also bring us emotional experiences. How do the emotional fluctuations brought by books affect readers’ evaluation of them? What is the difference in emotional fluctuations between books of different popularity? In this paper, we model and analyse the emotional fluctuations of different fiction books with different popularity and study the feasibility of predicting the popularity of fiction books using emotional fluctuations and recurrent neural networks. A new dataset is also generated to support this research and other related researches. Our proposed method obtained the best accuracy of 73.4% for predicting the popularity of fiction books and 41.4% for predicting genres. Some interesting data insights are also extracted from the dataset.

Authors:
Wang Xing, Zhang Shouhua, Smetannikov Ivan

Publication type:
A4 Article in conference proceedings

Place of publication:
CCRIS 2020: 2020 International Conference on Control, Robotics and Intelligent System

Keywords:
Emotion Analysis, natural language processing, popularity prediction

Published:

Full citation:
Xing Wang, Shouhua Zhang, and Ivan Smetannikov. 2020. Fiction Popularity Prediction Based on Emotion Analysis. In 2020 International Conference on Control, Robotics and Intelligent System (CCRIS 2020). Association for Computing Machinery, New York, NY, USA, 169–175. DOI:https://doi.org/10.1145/3437802.3437831

DOI:
https://doi.org/10.1145/3437802.3437831

Read the publication here:
http://urn.fi/urn:nbn:fi-fe202101262673