Related Experiment Video
Updated: Nov 8, 2025
04:17
DNA Virus Detection System Based on RPA-CRISPR/Cas12a-SPM and Deep Learning
Published on: May 10, 2024
1.1K
Identifying HIV-related digital social influencers using an iterative deep learning approach
Cheng Zheng1, Wei Wang1, Sean D Young2,3
1Department of Computer Science, University of California, Los Angeles.
AIDS (London, England)
|April 19, 2021
Summary
A deep learning model successfully identified 23 new online HIV influencers, improving HIV prevention campaign outreach. This AI approach efficiently discovers key voices in public health discussions.
Area of Science:
- Artificial Intelligence
- Public Health
- Social Network Analysis
Background:
- Community leaders are vital for HIV prevention interventions.
- Identifying online HIV influencers is challenging due to the dynamic nature of social media.
Purpose of the Study:
- To develop an iterative deep learning framework for automatically discovering HIV-related online social influencers.
- To address the difficulty in identifying qualified and willing influencers for HIV campaigns.
Main Methods:
- Utilized a deep learning graph neural network model on Twitter data (March 2018-March 2020) from users mentioning 'HIV' or 'AIDS'.
- Used two expert-identified 'online HIV influencers' as seed data for training.
- Modeled social influence to discover new potential influencers and validated results through manual verification.
Main Results:
- Identified 23 new, manually verified HIV-related influencers, including organizations and advocates.
- The proposed model demonstrated superior performance, achieving the highest accuracy/recall with an average improvement of 38.5% over baseline models.
Conclusions:
- Iterative deep learning models can effectively automate the identification of evolving key HIV influencers online.
- This approach offers significant potential for HIV researchers and departments to enhance prevention campaign reach within affected communities using big data.
