Effective Intra-Inter Interaction Learning for Relational Tables
This paper proposes Intra-Inter Relational Table Learning (InRTL), a unified framework that explicitly models dependencies both within and across relational tables, and formalizes two complementary interaction patterns: intra-table interactions and inter-table interactions.
Wei-Chen Li, Ken Zhong, Zheng Wang et al.
· Proceedings of the 32nd ACM... · 0 citations