On Poetic Topic Modeling: Extracting Themes and Motifs From a Corpus of Spanish Poetry

Abstract

This paper analyzes the application of LDA topic modeling to a corpus of poetry. First, it explains how the most coherent LDA-topics have been established by running several tests and automatically evaluating the coherence of the resulting LDA-topics. Results show, on one hand, that when dealing with a corpus of poetry, lemmatization is not advisable because several poetic features are lost in the process; and, on the other hand, that a standard LDA algorithm is better than a specific version of LDA for short texts (LF-LDA). The resulting LDA-topics have then been manually analyzed in order to define the relation between word topics and poems. The analysis shows that there are mainly two kinds of semantic relations: an LDA-topic could represent the subject or theme of the poem, but it could also represent a poetic motif. All these analyses have been undertaken on a large corpus of Golden Age Spanish sonnets. Finally, the paper shows the most relevant themes and motifs in this corpus such as "love," "religion," "heroics," "moral," or "mockery" on one hand, and "rhyme," "marine," "music," or "painting" on the other hand.

The Effects of Emotional Design on Multimedia Learning and Appreciation of Chinese Poetry

“…In the current literature, studies have focused more on the superficial textual characters of poetry, such as rhyme (Jacobs, 2015), and its influence on poetry appreciation. Rhyme of poems in different languages, such as Chinese (Chen et al, 2016;Chen and Yang, 2017), English (Xue et al, 2019), German (Obermeier et al, 2013;Lüdtke et al, 2014), and Spanish (Navarro-Colorado, 2018), has received extensive attention from researchers. However, less attention has been directed to the esthetic appreciation and in-depth learning of poetry in multimedia learning classrooms.…”

Topic models do not model topics: epistemological remarks and steps towards best practices

“…It was initially developed and continues to be used for information retrieval and text classification purposes in applied contexts (Bao et al, 2009;Asuncion et al, 2010;Ramage et al, 2011;Wang and Blei, 2011;Chuang et al, 2013;Si et al, 2014;Zhong et al, 2015;van Der Hooft et al, 2016;Liu et al, 2016;Boyd-Graber et al, 2017;Kuhn, 2018;Liu et al, 2019;Reber, 2019;Korfiatis et al, 2019, and many others). More recently, it has also gained momentum in the context of so-called distant reading 2 in the digital humanities and social sciences, where it is now increasingly being used to answer subject-specific research questions regarding the distributions of content in literary text (Asgari et al, 2013;Tangherlini and Leonard, 2013;Jockers and Mimno, 2013;Underwood, 2014;Goldstone and Underwood, 2014;Weitin and Herget, 2017;Mitrofanova and Sedova, 2017;Schöch, 2017;Erlin, 2018;Navarro-Colorado, 2018;Jacobs, 2018;Sieg, 2019;Dahllöf and Berglund, 2019;Liu and Jin, 2020), court decisions (Livermore et al, 2016;Panagis et al, 2016;Carter et al, 2016;Law, 2016;Wang et al, 2017;Rice, 2017;Young, 2019;Lampach and Dyevre, 2018), political and legal debate (Young, 2012;…”

Adaptive Mixed Component LDA for Low Resource Topic Modeling

“…We examine topic modeling in a low resource data setting (Hao et al, 2018), which has seen little attention but is commonly encountered in the digital humanities where document collections are potentially small (Jockers and Mimno, 2013;Schöch, 2017;Navarro-Colorado, 2018). 1 In such scenarios, word co-occurence statistics are less reliable due to sparsity of discrete counts.…”