<abstract xmlns="http://www.w3.org/1999/xhtml">

<p>This paper combines the LDA theme mining model with social network algorithms to construct an objective evaluation model for the imagery of cherry blossoms in Japanese literature and culture. To determine the number of topics, one needs to observe the change in perplexity degree for various numbers of topics. The degree of influence of surrounding nodes on this node is measured using in-degree and out-degree, and the degree of closeness between users in social networks is measured using the clustering coefficient. Based on the network data, we analyzed Japanese literary works related to cherry blossoms and readers’ comments on related literary review websites, followed by analyzing the rate of microblog retweets when a hot literary work was published to determine the theme density. In the beginning, Topic 1, which represents positivity, has the highest theme intensity of 0.2581, indicating that the imagery of cherry blossoms is skewed toward the positive in the minds of readers and netizens. Following the emergence of hot literature, the spread of Topic2, which is associated with negativity, reached its peak, with a density that reached as high as 2235721 times. At this time, the highest high-frequency words are death 5827 times, and poignant 5748 times, all of which are negative meanings, and the theme intensity of Topic 2 is 0.2992. The impression of people can be negatively impacted by cherry blossoms. By constructing a model for literary analysis of imagery, this paper serves as a reference and presents a new direction for textual research.</p>
</abstract>

This paper combines the LDA theme mining model with social network algorithms to construct an objective evaluation model for the imagery of cherry blossoms in Japanese literature and culture. To determine the number of topics, one needs to observe the change in perplexity degree for various numbers of topics. The degree of influence of surrounding nodes on this node is measured using in-degree and out-degree, and the degree of closeness between users in social networks is measured using the clustering coefficient. Based on the network data, we analyzed Japanese literary works related to cherry blossoms and readers’ comments on related literary review websites, followed by analyzing the rate of microblog retweets when a hot literary work was published to determine the theme density. In the beginning, Topic 1, which represents positivity, has the highest theme intensity of 0.2581, indicating that the imagery of cherry blossoms is skewed toward the positive in the minds of readers and netizens. Following the emergence of hot literature, the spread of Topic2, which is associated with negativity, reached its peak, with a density that reached as high as 2235721 times. At this time, the highest high-frequency words are death 5827 times, and poignant 5748 times, all of which are negative meanings, and the theme intensity of Topic 2 is 0.2992. The impression of people can be negatively impacted by cherry blossoms. By constructing a model for literary analysis of imagery, this paper serves as a reference and presents a new direction for textual research.

Exploring the Flux of Cherry Blossom Imagery in Japanese Literature and Culture by Combining Social Network Analysis Methods

Applied Mathematics and Nonlinear Sciences

This work is licensed under the Creative Commons Attribution 4.0 International License.

This paper combines the LDA theme mining model with social network algorithms to construct an objective evaluation model for the imagery of cherry blossoms in...

Exploring the Flux of Cherry Blossom Imagery in Japanese Literature and Culture by Combining Social Network Analysis Methods

Data publikacji: 31 sty 2024

Zakres stron: -

Otrzymano: 26 gru 2023

Przyjęty: 31 gru 2023

DOI: https://doi.org/10.2478/amns-2024-0304

Słowa kluczowe
LDA topic mining, Perplexity, Social network algorithm, Topic density, Cherry blossom imagery

© 2024 Meizi Han, published by Sciendo

This work is licensed under the Creative Commons Attribution 4.0 International License.

Exploring the Flux of Cherry Blossom Imagery in Japanese Literature and Culture by Combining Social Network Analysis Methods

Data publikacji: 31 sty 2024

Zakres stron: -

Otrzymano: 26 gru 2023

Przyjęty: 31 gru 2023

DOI: https://doi.org/10.2478/amns-2024-0304

Słowa kluczoweLDA topic mining, Perplexity, Social network algorithm, Topic density, Cherry blossom imagery

© 2024 Meizi Han, published by Sciendo

This work is licensed under the Creative Commons Attribution 4.0 International License.

Słowa kluczowe
LDA topic mining, Perplexity, Social network algorithm, Topic density, Cherry blossom imagery