<![CDATA[This research investigates societal perspectives on the childfree lifestyle through Intent Sentiment Analysis, combining Latent Dirichlet Allocation (LDA) and Support Vector Machine (SVM) techniques. The childfree lifestyle, a deliberate decision by individuals or couples to remain childless, has spurred extensive public discourse, particularly on platforms like Twitter. This research aims to analyze sentiments and intentions within these discussions to uncover their implications for social dynamics and familial relationships. Using LDA, dominant topics were identified from a dataset of Twitter comments on the childfree topic. LDA uncovered hidden themes by modeling topics as mixtures of words, which were subsequently classified into positive, negative, and neutral sentiments using SVM. Data preprocessing included cleaning, tokenization, and stop word removal, while oversampling with SMOTE addressed class imbalances. The optimal number of topics was determined using coherence scores, with the highest coherence value of 0.400 achieved at one topic. The findings revealed that positive sentiments were classified more effectively than negative and neutral sentiments when using LDA and SVM with SMOTE. The top 10 topics primarily reflected societal commentary on the childfree lifestyle. Challenges included incomplete preprocessing, suboptimal clustering of similar themes, and imbalanced data, which limited the effectiveness of topic modeling and classification. Addressing these issues through improved feature selection, parameter optimization, and data augmentation could enhance performance for underrepresented categories. This research provides valuable insights into public attitudes toward the childfree lifestyle, offering implications for social research and policy development in the context of evolving societal norms. ]]>
Copyrights © 2025
