Multi-task Learning over Graph Structures

2018/11/26 by Pengfei Liu, Jie Fu, Liu, Pengfei +7
Computer Science · #Computation and Language (cs.CL) #FOS: Computer and information sciences #cs.CL

paper · pdf · doi:10.48550/arxiv.1811.10211

Accepted by AAAI 2019

arxiv created 2018/11/26 · arxiv updated 2018/11/27

Abstract

We present two architectures for multi-task learning with neural sequence models. Our approach allows the relationships between different tasks to be learned dynamically, rather than using an ad-hoc pre-defined structure as in previous work. We adopt the idea from message-passing graph neural networks and propose a general graph multi-task learning framework in which different tasks can communicate with each other in an effective and interpretable way. We conduct extensive experiments in text classification and sequence labeling to evaluate our approach on multi-task learning and transfer learning. The empirical results show that our models not only outperform competitive baselines but also learn interpretable and transferable patterns across tasks.

Related

• Learning an Interpretable Graph Structure in Multi-Task Learning
• A Meta-Learning Approach for Graph Representation Learning in Multi-Task Settings
• All in One: Multi-Task Prompting for Graph Neural Networks (Extended Abstract)
• A Generalized Recurrent Neural Architecture for Text Classification with Multi-Task Learning
• Recurrent Neural Network for Text Classification with Multi-Task Learning