The Study of Poetical Style Attribution and Classification of Poetic Subgenres in the Tang and Song Poetry

Abstract

The literary styles of Tang and Song poetry exhibit noticeable differences, at times directly denoting two distinct categories of poetic styles. However, sometimes the poetic style of Tang Dynasty poets may be more akin to Song poetry, and the poetic style of Song Dynasty poets may lean closer to that of the Tang Dynasty. This study employs quantitative analysis and establishes mathematical models to investigate these differences. Methods: Firstly, this paper employs a logistic regression model based on the simulated annealing algorithm to classify the style of poets and determine their style affiliation. Secondly, using the k-means clustering model, Tang and Song poetic styles are further refined into subcategories. Finally, scores are computed using the TOPSIS model modified by the entropy weight method to select the most representative poems and poets within each style. Conclusions: 1) The model identifies the style affiliation of Pei Che and Liu Yizhi as Song poetry style and Tang poetry style, respectively, with an accuracy rate of 83.3%. 2) The Tang poetic style is divided into three categories, and the Song poetic style is divided into five categories, with the first subcategory of Tang poetry including poems like "Passing Jin Yang Palace" and "Spring Platform Views." 3) The most representative poems include "Sending Xue Shaoqing to Qingyang" and "Introducing the Ballad for the Feast."