Research map / Large language models / AI and society
Understanding Microtargeting Pattern on Social Media
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's full text. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
A research-statement overview summarizing the author's computational social science and NLP work on characterizing user types, messaging, and themes/arguments in microtargeted social media content.
Problem
Understanding how microtargeted messaging on social media varies across demographic and interest groups, and how it affects users (transparency, polarization, harmful messaging).
Approach
Frames microtargeting research as a combination of user characterization, messaging analysis, and theme/argument understanding, using graph embeddings, weak supervision, and LLM-based machine-in-the-loop methods on political, climate, and COVID-19 vaccine ads.
Key results
- Pro-energy climate ads receive the most views in Texas, while clean-energy ads are most viewed in California (as reported from prior work).
- COVID-19 vaccine ads from the same source shift moral foundations (liberty/oppression vs care/harm) depending on targeted demographic (older population vs women of reproductive age).
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| target older users rather than women of reproductive age | COVID-19 vaccine campaigns · COVID-19 vaccine campaigns | message themes | Mixed |
| use pro-energy rather than clean-energy ad messaging | climate advertising · Texas and California | views | Mixed |
Limitations
- Summarizes prior work rather than presenting new experiments or results
- No detailed methodology or evaluation reported in this statement
Experimental setup
Datasets: political ads (2020 U.S. presidential election) climate advertising campaigns COVID-19 vaccine campaign posts
Closest papers by meaning
- Graph Bayesian Optimization for Multiplex Influence Maximization (AAAI 2024 · similarity 0.78)
- Private Blotto: Viewpoint Competition with Polarized Agents (AAAI 2025 · similarity 0.78)
- A Persuasive Approach to Combating Misinformation (ICML 2024 · similarity 0.78)
- MIDDAG: Where Does Our News Go? Investigating Information Diffusion via Community-Level Information Pathways (AAAI 2024 · similarity 0.77)
- Exploiting Fine-Grained Skip Behaviors for Micro-Video Recommendation (AAAI 2025 · similarity 0.77)
- SMTPD: A New Benchmark for Temporal Prediction of Social Media Popularity (CVPR 2025 · similarity 0.77)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.