Word Clouds Based on Topic-Modeling Per Century!

This study embarked on an exploration of poetry themes across four centuries, utilizing computational techniques to reveal insights into the predominant topics of poetic discourse. The methodology involved gathering a collection of poems from each century, meticulously processing them to extract textual data. Each century's corpus underwent preprocessing steps where common stopwords were removed, ensuring that the subsequent analysis focused on meaningful content words.

​

Topic Modeling techniques, specifically Latent Dirichlet Allocation (LDA), were employed to uncover underlying themes within the poetry. LDA facilitated the identification of distinct topics by modeling how words co-occur in the poems. For each century, the model discerned several topics, each characterized by a set of key terms that encapsulated the prevalent subjects and motifs in the poetry.

​

Word Cloud Generation was utilized as a visual representation tool to portray the essence of each identified topic. Word clouds succinctly depicted the frequency of top terms associated with each topic, with larger font sizes indicating higher frequency. These visualizations offered a clear glimpse into the thematic landscapes of poetry from different centuries, showcasing how poetic discourse evolved and persisted across time.