Reference

[apple2373]

It is not mentioned, but I am wondering this repository is the implementation of this paper: http://www.kdd.org/kdd2016/subtopic/view/node2vec-scalable-feature-learning-for-networks

Is that right?

[palmagro]

Hi Satoshi,

No, thats not the implementation of this paper. We are researchers in
Universidad Central de Ecuador and Universidad de Sevilla working in
Machine Learaning over Multi-Relational Graphs. We are working in our own
research and by coincidence we used the same name: node2vec.

Our system is quite different of this one presented in the paper you
referenced, it uses neo4j as storage system, uses properties (in addition
of connections) to perform the embedding and the machine learning algorithm
to use to learn from the embedding is a free parameter.

Apologies for the state of the repo, we will refactor it and include a
complete documentation soon.

Regards.

Pedro Almagro

2016-05-28 0:58 GMT+02:00 Satoshi Tsutsui notifications@github.com:

It is not mentioned, but I am wondering this repository is the
implementation of this paper:
http://www.kdd.org/kdd2016/subtopic/view/node2vec-scalable-feature-learning-for-networks

Is that right?

—
You are receiving this because you are subscribed to this thread.
Reply to this email directly, view it on GitHub
#1, or mute the thread
https://github.com/notifications/unsubscribe/ABDKXDRuwWYiuZCn5IgSbkuPeOSzbwxcks5qF3ctgaJpZM4Io7FN
.

[apple2373]

Sure. Thank you for clarification. I am investigating graph embedding methods. Looking forward to the documentation soon! --Satoshi