Social Influence in Networks

Interactions among streamers and viewers on a live-streaming platform

Online behavior is shaped by interaction. My research studies how streamer–viewer exchanges—such as donations and responses—and viewer–viewer relationships influence interests, engagement, and content over time.

By modeling these feedback loops, we can understand the co-evolution of communities and content, then build recommendations that respond to changing social dynamics rather than treating users as isolated individuals.

Keywords

Social NetworksUser InteractionLive StreamingCo-evolution

Selected Publications

Learning the Co-evolution Process on Live Stream Platforms with Dual Self-attention for Next-topic Recommendations 2023 · CIKM

Live Multi-Streaming and Donation Recommendations via Coupled Donation-Response Tensor Factorization 2020 · CIKM

Social Attentive Network for Live Stream Recommendation 2020 · WWW