88. A New Approach to Sentiment Analysis in Spanish Language Social Media
Invited abstract in session MC-4: Analytics for Mis/Disinformation in Healthcare, stream Regular talks.
Monday, 11:00-12:30Room: Room S3
Authors (first author is the speaker)
| 1. | Fernando Wilson
|
| University of Utah |
Abstract
This study aims to develop and validate a Spanish version of the Valence Aware Dictionary and sEntiment Reasoner (S-VADER) for analyzing sentiment in Spanish language social media posts. The goal is to improve communication strategies and outcomes among Spanish-speaking populations by monitoring and mitigating mis- and disinformation concerning health or healthcare-related information such as eligibility for health benefits programs. A sentiment-labeled lexicon is developed using a crowd-sourced platform (Prolific), analyzing 5,000 randomly sampled posts from Facebook/Instagram and Reddit. Native Spanish-speaking raters, based in the US or Spanish-speaking countries, will evaluate these lexical features on a scale of -4 to +4, denoting direction and intensity of sentiment. Raters will undergo screening for Spanish fluency and will be provided with a standardized set of criteria to ensure consistency in sentiment assessment. The validated S-VADER tool will be used to perform sentiment analysis on health-related social media content, helping identify the proliferation and impact of health mis- or disinformation across Spanish-speaking communities. The S-VADER tool can be applied in a number of ways to design and optimize targeted public health interventions and communication strategies that are culturally and linguistically tailored, including whether messaging is being accepted or rejected by Spanish speakers and the intensity of reactions to the spread of false information.
Keywords
- Artificial Intelligence
- Analytics
- Decision support
Status: accepted
Back to the list of papers