search

gasteigerjo / ppnp

PPNP & APPNP models from "Predict then Propagate: Graph Neural Networks meet Personalized PageRank" (ICLR 2019)
https://www.daml.in.tum.de/ppnp
MIT License
318 stars 53 forks source link

About comparing to Ying. 2018 #2

Closed yzthz closed 5 years ago

yzthz commented 5 years ago

Hi,

Thanks for the nice work! wonder if you could give a short description of the difference between this work's message passing and that of Ying. 2018, Graph Convolutional Neural Networks for Web-Scale Recommender Systems? Both use (approximate) personalized page rank for message passing. Is there any procedural difference?

Thanks! D

gasteigerjo commented 5 years ago

Thank you for the interest in our work! First of all, Ying's PinSage model does not use personalized PageRank. PinSage uses (explicit) random walks, which approximate regular PageRank in the limit of infinite length. It therefore exhibits all the problems described in our paper.

Additionally, PPNP is not a regular message passing neural network (MPNN). Instead, PPNP disentangles prediction and propagation by first using a neural network for prediction and then propagating these predictions using PPR.