基于自然语言处理的文学文本主题分析

Analyzing literary topics: An NLP approach

本研究使用R的mallet程序包对自建的英汉散文可比较语料库进行主题建模,并利用可视化方式对结果进行详细的对比分析。研究发现,英汉语料库之间既有相同的重要主题,同时也存在一些有意义的差异。在方法论上,本研究验证了主题建模在文学文本分析(包括跨语言对比分析)中的应用价值。

This study attempts an automatic analysis of latent topics in our self-constructed corpora of English and Chinese essays in the light of Digital Humanities,by using the R package mallet and a variety of visualization techniques.It is discovered that the selected English and Chinese essays have both important common topics and meaningful differences.Methodology-wise,it is also illustrated that topic modeling has significant potentials in the analyses of literary texts.