# Estimating Online Influence Needs Causal Modeling! Counterfactual Analysis of Social Media Engagement

arXiv: https://arxiv.org/abs/2505.19355 · HTML: https://mlsanity.com/paper/2505.19355/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 22 of its 52 references that are in the map.
- First posted on arXiv 25 May 2025.

## Builds on (most cited first, 22 of 22)

- [Attention Is All You Need](https://mlsanity.com/paper/1706.03762/) (2017)
- [Mamba: Linear-Time Sequence Modeling with Selective State Spaces](https://mlsanity.com/paper/2312.00752/) (2023)
- [Parameter-Efficient Transfer Learning for NLP](https://mlsanity.com/paper/1902.00751/) (2019)
- [FlashAttention: Fast and Memory-Efficient Exact Attention with IO-Awareness](https://mlsanity.com/paper/2205.14135/) (2022)
- [Informer: Beyond Efficient Transformer for Long Sequence Time-Series Forecasting](https://mlsanity.com/paper/2012.07436/) (2020)
- [Autoformer: Decomposition Transformers with Auto-Correlation for Long-Term Series Forecasting](https://mlsanity.com/paper/2106.13008/) (2021)
- [HiPPO: Recurrent Memory with Optimal Polynomial Projections](https://mlsanity.com/paper/2008.07669/) (2020)
- [Transformer Hawkes Process](https://mlsanity.com/paper/2002.09291/) (2020)
- [DeepInf: Social Influence Prediction with Deep Learning](https://mlsanity.com/paper/1807.05560/) (2018)
- [SEISMIC: A Self-Exciting Point Process Model for Predicting Tweet Popularity](https://mlsanity.com/paper/1506.02594/) (2015)
- [DeepCas: an End-to-end Predictor of Information Cascades](https://mlsanity.com/paper/1611.05373/) (2016)
- [Can Cascades be Predicted?](https://mlsanity.com/paper/1403.4608/) (2014)
- [Expecting to be HIP: Hawkes Intensity Processes for Social Media Popularity](https://mlsanity.com/paper/1602.06033/) (2016)
- [A Survey of Information Cascade Analysis: Models, Predictions, and Recent Advances](https://mlsanity.com/paper/2005.11041/) (2020)
- [Topological Recurrent Neural Network for Diffusion Prediction](https://mlsanity.com/paper/1711.10162/) (2017)
- [Interval-censored Transformer Hawkes: Detecting Information Operations using the Reaction of Social Systems](https://mlsanity.com/paper/2211.14114/) (2022)
- [Slipping to the Extreme: A Mixed Method to Explain How Extreme Opinions Infiltrate Online Discussions](https://mlsanity.com/paper/2109.00302/) (2021)
- [Continuous-Time Graph Learning for Cascade Popularity Prediction](https://mlsanity.com/paper/2306.03756/) (2023)
- [Before It's Too Late: A State Space Model for the Early Prediction of Misinformation and Disinformation Engagement](https://mlsanity.com/paper/2502.04655/) (2025)
- [Online Popularity under Promotion: Viral Potential, Forecasting, and the Economics of Time](https://mlsanity.com/paper/1703.01012/) (2017)
- [Interval-censored Hawkes processes](https://mlsanity.com/paper/2104.07932/) (2021)
- [Opinion Market Model: Stemming Far-Right Opinion Spread using Positive Interventions](https://mlsanity.com/paper/2208.06620/) (2022)

## Built on by (most cited first, 1 of 1)

- [Mario at EXIST 2025: A Simple Gateway to Effective Multilingual Sexism Detection](https://mlsanity.com/paper/2507.10996/) (2025)

## Abstract (arXiv)

Understanding true influence in social media requires distinguishing correlation from causation--particularly when analyzing misinformation spread. While existing approaches focus on exposure metrics and network structures, they often fail to capture the causal mechanisms by which external temporal signals trigger engagement. We introduce a novel joint treatment-outcome framework that leverages existing sequential models to simultaneously adapt to both policy timing and engagement effects. Our approach adapts causal inference techniques from healthcare to estimate Average Treatment Effects (ATE) within the sequential nature of social media interactions, tackling challenges from external confounding signals. Through our experiments on real-world misinformation and disinformation datasets, we show that our models outperform existing benchmarks by 15--22% in predicting engagement across diverse counterfactual scenarios, including exposure adjustment, timing shifts, and varied intervention durations. Case studies on 492 social media users show our causal effect measure aligns strongly with the gold standard in influence estimation, the expert-based empirical influence.