The use of artificial intelligence and natural language processing for visualizing social media discussions surrounding the repeal of Roe v. Wade in the United States of America
More details
Hide details
Central Michigan University, 122 Falling Rock Way, United States
Central Michigan University, 1200 S. Franklin St., Mount Pleasant, Mich. 48859, United States
Publication date: 2023-04-26
Popul. Med. 2023;5(Supplement):A1860
Background and Objective:
We present a novel method of network visualization with computer-assisted processing of text. We leverage human and computer collaboration to develop and validate algorithms. We apply a network analysis and natural language processing to the corpus of 28,693 YouTube comments surrounding the repeal of Roe v. Wade. In addition, we connect visualizations to qualitative study, to deepen our exploration of social media users’ understanding of policy and implications for women’s health.

We analyzed comments to nine most watched YouTube newscasts about the repeal of Roe v. Wade by ABC, CBS, CNN, Fox News, MSNC and Vice News. We applied a Gender API through google and computer-scored social and psychological states from Linguistic Inquiry Word Count (LIWC-22). VOSviewer visualizations of Roe v. Wade relevant terms extracted from YouTube comments are overlayed with computer-generated LIWC scores.

Our colorful map – a network of 256 interconnected terms extracted from 28,693 comments has 4 thematic clusters of terms: 1) overturning of Roe v. Wade/states’ rights/political debate/Supreme Court rulings/abortion rights; 2) Consequences of eliminating the right to choose/possible solutions; and 3) Argument for determining when life begins/fetal development/health complications; 4) Religion and morality. Clusters 2 and 3 score high on female usernames (Gender API), in-text references to females (LIWC), and health related concerns. In contrast, male usernames are more prevalent in Clusters 1 and 4 with discussions on religion, policy, and the political implications of the Supreme Court decision.

Eye-catching term networks and LIWC overlays help researchers visualize and communicate about policy issues, as they are deliberated in social media. Overlayed with gender-specific information, networks of co-occurring terms can assist in audience segmentation. To make data analysis more efficient, some human validation and curation is needed with automated text scoring to identify discussion areas with specific linguistic features.

Journals System - logo
Scroll to top