Sentiment Dynamics of The Chronicles of Narnia and Their Ranking.

Kaiyun Dai,Menglan Ma,Jianbo Gao
DOI: https://doi.org/10.1007/978-3-319-93372-6_24
2018-01-01
Abstract:Everyone in a civilized society grows up by reading stories. Fictions, including those for children, are an important type of stories, as they reflect social and cultural reality to some degree. The plot, the figures, and the environment of a fiction are the three main elements of a fiction. In particular, the development of the plot is pivotal for a fiction to be successful. It is now generally thought that sentiment dynamics of the fiction can well reflect the plot development. With the availability of a number of algorithms to automatically obtain the sentiment dynamics of a fiction, it has become increasingly desirable to fully understand its sentiment dynamics. This motivates us to use random fractal theory to study a set of popular children's fictions, The Chronicles of Narnia, written by the famed author, C.S. Lewis. We find the sentiment dynamics of each novel of the series possesses persistent long-range correlations, characterized by a Hurst parameter larger than 1/2. This has offered a mechanism to understand why many sentiment dynamics occurring naturally in a society or imagined by an author of a fiction can arouse strong emotions in humans. Interestingly, the value of the Hurst parameter for the series is strongly positively correlated with the score of the novels from Goodreads, suggesting that the scaling law governing sentiment dynamics can be used to objectively appraise the optimality of a fiction.
What problem does this paper attempt to address?