LiGNN: Graph Neural Networks at LinkedIn
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , , , , , , , , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2024
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866911778860957696 |
|---|---|
| author | Borisyuk, Fedor He, Shihai Ouyang, Yunbo Ramezani, Morteza Du, Peng Hou, Xiaochen Jiang, Chengming Pasumarthy, Nitin Bannur, Priya Tiwana, Birjodh Liu, Ping Dangi, Siddharth Sun, Daqi Pei, Zhoutao Shi, Xiao Zhu, Sirou Shen, Qianqi Lee, Kuang-Hsuan Stein, David Li, Baolei Wei, Haichao Ghoting, Amol Ghosh, Souvik |
| author_facet | Borisyuk, Fedor He, Shihai Ouyang, Yunbo Ramezani, Morteza Du, Peng Hou, Xiaochen Jiang, Chengming Pasumarthy, Nitin Bannur, Priya Tiwana, Birjodh Liu, Ping Dangi, Siddharth Sun, Daqi Pei, Zhoutao Shi, Xiao Zhu, Sirou Shen, Qianqi Lee, Kuang-Hsuan Stein, David Li, Baolei Wei, Haichao Ghoting, Amol Ghosh, Souvik |
| contents | In this paper, we present LiGNN, a deployed large-scale Graph Neural Networks (GNNs) Framework. We share our insight on developing and deployment of GNNs at large scale at LinkedIn. We present a set of algorithmic improvements to the quality of GNN representation learning including temporal graph architectures with long term losses, effective cold start solutions via graph densification, ID embeddings and multi-hop neighbor sampling. We explain how we built and sped up by 7x our large-scale training on LinkedIn graphs with adaptive sampling of neighbors, grouping and slicing of training data batches, specialized shared-memory queue and local gradient optimization. We summarize our deployment lessons and learnings gathered from A/B test experiments. The techniques presented in this work have contributed to an approximate relative improvements of 1% of Job application hearing back rate, 2% Ads CTR lift, 0.5% of Feed engaged daily active users, 0.2% session lift and 0.1% weekly active user lift from people recommendation. We believe that this work can provide practical solutions and insights for engineers who are interested in applying Graph neural networks at large scale. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2402_11139 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | LiGNN: Graph Neural Networks at LinkedIn Borisyuk, Fedor He, Shihai Ouyang, Yunbo Ramezani, Morteza Du, Peng Hou, Xiaochen Jiang, Chengming Pasumarthy, Nitin Bannur, Priya Tiwana, Birjodh Liu, Ping Dangi, Siddharth Sun, Daqi Pei, Zhoutao Shi, Xiao Zhu, Sirou Shen, Qianqi Lee, Kuang-Hsuan Stein, David Li, Baolei Wei, Haichao Ghoting, Amol Ghosh, Souvik Machine Learning Artificial Intelligence In this paper, we present LiGNN, a deployed large-scale Graph Neural Networks (GNNs) Framework. We share our insight on developing and deployment of GNNs at large scale at LinkedIn. We present a set of algorithmic improvements to the quality of GNN representation learning including temporal graph architectures with long term losses, effective cold start solutions via graph densification, ID embeddings and multi-hop neighbor sampling. We explain how we built and sped up by 7x our large-scale training on LinkedIn graphs with adaptive sampling of neighbors, grouping and slicing of training data batches, specialized shared-memory queue and local gradient optimization. We summarize our deployment lessons and learnings gathered from A/B test experiments. The techniques presented in this work have contributed to an approximate relative improvements of 1% of Job application hearing back rate, 2% Ads CTR lift, 0.5% of Feed engaged daily active users, 0.2% session lift and 0.1% weekly active user lift from people recommendation. We believe that this work can provide practical solutions and insights for engineers who are interested in applying Graph neural networks at large scale. |
| title | LiGNN: Graph Neural Networks at LinkedIn |
| topic | Machine Learning Artificial Intelligence |
| url | https://arxiv.org/abs/2402.11139 |