MPDRL

All author and affiliation information has been removed from our source code.

Overview

This is a demo of Meta-path Discovery with Reinforcement Learning (MPDRL)

Environment

1. python=3.6.3

2. pytorch=0.4.1

Data

• Yago CoreFacts dataset is available at Yago Downloads

• NELL-1,555-th dump is available at Never-ending Language Learning-