# Entity Representation Learning Through Onsite-Offsite Graph for Pinterest Ads

arXiv: https://arxiv.org/abs/2508.02609 · HTML: https://mlsanity.com/paper/2508.02609/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 8 of its 17 references that are in the map.
- First posted on arXiv 4 Aug 2025.

## Builds on (most cited first, 8 of 8)

- [Inductive Representation Learning on Large Graphs](https://mlsanity.com/paper/1706.02216/) (2017)
- [PyTorch-BigGraph: A Large-scale Graph Embedding System](https://mlsanity.com/paper/1903.12287/) (2019)
- [SLAPS: Self-Supervision Improves Structure Learning for Graph Neural Networks](https://mlsanity.com/paper/2102.05034/) (2021)
- [LLMs as Zero-shot Graph Learners: Alignment of GNN Representations with LLM Token Embeddings](https://mlsanity.com/paper/2408.14512/) (2024)
- [Grale: Designing Networks for Graph Learning](https://mlsanity.com/paper/2007.12002/) (2020)
- [ItemSage: Learning Product Embeddings for Shopping Recommendations at Pinterest](https://mlsanity.com/paper/2205.11728/) (2022)
- [M2GRL: A Multi-task Multi-view Graph Representation Learning Framework for Web-scale Recommender Systems](https://mlsanity.com/paper/2005.10110/) (2020)
- [Heterogeneous Graph Neural Networks for Large-Scale Bid Keyword Matching](https://mlsanity.com/paper/2111.00926/) (2021)

## Abstract (arXiv)

Graph Neural Networks (GNN) have been extensively applied to industry recommendation systems, as seen in models like GraphSageciteGraphSage, TwHIMciteTwHIM, LiGNNciteLiGNN etc. In these works, graphs were constructed based on users' activities on the platforms, and various graph models were developed to effectively learn node embeddings. In addition to users' onsite activities, their offsite conversions are crucial for Ads models to capture their shopping interest. To better leverage offsite conversion data and explore the connection between onsite and offsite activities, we constructed a large-scale heterogeneous graph based on users' onsite ad interactions and opt-in offsite conversion activities. Furthermore, we introduced TransRA (TransRciteTransR with Anchors), a novel Knowledge Graph Embedding (KGE) model, to more efficiently integrate graph embeddings into Ads ranking models. However, our Ads ranking models initially struggled to directly incorporate Knowledge Graph Embeddings (KGE), and only modest gains were observed during offline experiments. To address this challenge, we employed the Large ID Embedding Table technique and innovated an attention based KGE finetuning approach within the Ads ranking models. As a result, we observed a significant AUC lift in Click-Through Rate (CTR) and Conversion Rate (CVR) prediction models. Moreover, this framework has been deployed in Pinterest's Ads Engagement Model and contributed to 2.69% CTR lift and 1.34% CPC reduction. We believe the techniques presented in this paper can be leveraged by other large-scale industrial models.