Engineer friends often ask me: Graph Deep Learning sounds great, but are there any big commercial success stories? Is it being deployed in practical applications? Besides the obvious ones–recommendation systems at Pinterest, Alibaba and Twitter–a slightly nuanced success story is the Transformer architecture, which has taken the NLP industry by storm. Through this post, I want to establish links between Graph Neural Networks (GNNs) and Transformers. I’ll talk about the intuitions behind model architectures in the NLP and GNN communities, make connections using equations and figures, and discuss how we could work together to drive progress.
D. Krompass, M. Nickel, and V. Tresp. International Conference on Data Science and Advanced Analytics, DSAA 2014, Shanghai, China, October 30 - November 1, 2014, page 18--24. IEEE, (2014)
X. Xu, N. Yuruk, Z. Feng, and T. Schweiger. KDD '07: Proceedings of the 13th ACM SIGKDD international conference on Knowledge discovery and data mining, page 824--833. New York, NY, USA, ACM, (2007)
R. Bekkerman, R. El-Yaniv, and A. McCallum. ICML '05: Proceedings of the 22nd international conference on Machine learning, page 41--48. New York, NY, USA, ACM Press, (2005)
C. Tantipathananandh, T. Berger-Wolf, and D. Kempe. KDD '07: Proceedings of the 13th ACM SIGKDD international conference on Knowledge discovery and data mining, page 717--726. New York, NY, USA, ACM, (2007)
C. Schmitz, A. Hotho, R. Jäschke, and G. Stumme. Proceedings of the 3rd European Semantic Web Conference, volume 4011 of LNCS, page 530-544. Budva, Montenegro, Springer, (June 2006)