TwHIN: Embedding the Twitter Heterogeneous Information Network for Personalized Recommendation

Abstract

Social networks, such as Twitter, form a heterogeneous information network (HIN) where nodes represent domain entities (e.g., user, content, advertiser, etc.) and edges represent one of many entity interactions (e.g, a user re-sharing content or "following" another). Interactions from multiple relation types can encode valuable information about social network entities not fully captured by a single relation; for instance, a user's preference for accounts to follow may depend on both user-content engagement interactions and the other users they follow. In this work, we investigate knowledge-graph embeddings for entities in the Twitter HIN (TwHIN); we show that these pretrained representations yield significant offline and online improvement for a diverse range of downstream recommendation and classification tasks: personalized ads rankings, account follow-recommendation, offensive content detection, and search ranking. We discuss design choices and practical challenges of deploying industry-scale HIN embeddings, including compressing them to reduce end-to-end model latency and handling parameter drift across versions.

References in corpus (4)

• LINE: Large-scale Information Network Embedding
• DCN V2: Improved Deep & Cross Network and Practical Lessons for Web-scale Learning to Rank Systems
• PinnerSage: Multi-Modal User Embedding Framework for Recommendations at Pinterest
• An Analysis Of Entire Space Multi-Task Models For Post-Click Conversion Prediction

Cited by in corpus (7)

• Predicting Information Pathways Across Online Communities
• Scaling User Modeling: Large-scale Online User Representations for Ads Personalization in Meta
• A Scalable and Efficient Signal Integration System for Job Matching
• The Fallacy of Borda Count Method -- Why it is Useless with Group Intelligence and Shouldn't be Used with Big Data including Banking Customer Services
• Understanding the Design Principles of Link Prediction in Directed Settings
• kNN-Embed: Locally Smoothed Embedding Mixtures For Multi-interest Candidate Retrieval
• Blending Sequential Embeddings, Graphs, and Engineered Features: 4th Place Solution in RecSys Challenge 2025