Leveraging Prompt-Based Large Language Models: Predicting Pandemic Health Decisions and Outcomes Through Social Media Language
| dc.contributor.author | Ding, Xiaohan | en |
| dc.contributor.author | Carik, Buse | en |
| dc.contributor.author | Gunturi, Uma Sushmitha | en |
| dc.contributor.author | Reyna, Valerie | en |
| dc.contributor.author | Rho, Eugenia | en |
| dc.date.accessioned | 2024-06-04T18:48:31Z | en |
| dc.date.available | 2024-06-04T18:48:31Z | en |
| dc.date.issued | 2024-05-11 | en |
| dc.date.updated | 2024-06-01T08:00:20Z | en |
| dc.description.abstract | We introduce a multi-step reasoning framework using prompt-based LLMs to examine the relationship between social media lan guage patterns and trends in national health outcomes. Grounded in fuzzy-trace theory, which emphasizes the importance of “gists” of causal coherence in effective health communication, we introduce Role-Based Incremental Coaching (RBIC), a prompt-based LLM framework, to identify gists at-scale. Using RBIC, we systematically extract gists from subreddit discussions opposing COVID-19 health measures (Study 1). We then track how these gists evolve across key events (Study 2) and assess their influence on online engage ment (Study 3). Finally, we investigate how the volume of gists is associated with national health trends like vaccine uptake and hospitalizations (Study 4). Our work is the first to empirically link social media linguistic patterns to real-world public health trends, highlighting the potential of prompt-based LLMs in identifying critical online discussion patterns that can form the basis of public health communication strategies. | en |
| dc.description.version | Published version | en |
| dc.format.mimetype | application/pdf | en |
| dc.identifier.doi | https://doi.org/10.1145/3613904.3642117 | en |
| dc.identifier.uri | https://hdl.handle.net/10919/119262 | en |
| dc.language.iso | en | en |
| dc.publisher | ACM | en |
| dc.rights | Creative Commons Attribution 4.0 International | en |
| dc.rights.holder | The author(s) | en |
| dc.rights.uri | http://creativecommons.org/licenses/by/4.0/ | en |
| dc.title | Leveraging Prompt-Based Large Language Models: Predicting Pandemic Health Decisions and Outcomes Through Social Media Language | en |
| dc.type | Article - Refereed | en |
| dc.type.dcmitype | Text | en |