MLRG Presents at AAAI 2023

Feb 12, 2023. MLRG members Akshay Aravamudan and Xi Zhang attended the 37th AAAI Conference on Artificial Intelligence (AAAI 2023), held in Washington, DC, during February 7–14, 2023. At the conference, Aravamudan presented the group’s work titled “Anytime User Engagement Prediction in Information Cascades for Arbitrary Observation Periods.”

The presented work studies how to predict user engagement in information cascades, such as whether users will react to, share, or otherwise participate in the spread of online content. The paper introduces DANTE, a discriminative probabilistic model based on split-population multivariate survival processes. Unlike approaches that require separate models for different observation periods or forecast horizons, DANTE provides a single interpretable framework for making anytime predictions from partially observed cascades. This is significant because such forecasts can support applications in social-media analytics, online marketing, and misinformation mitigation, while also providing principled uncertainty estimates for predicted engagement counts.

Both Aravamudan and Zhang received travel scholarships from AAAI to attend the conference. Congratulations to Akshay and Xi on representing our group at AAAI 2023!