First major release

[Open Graph Benchmark]

Dear all,

We are excited to make our first major release of OGB! We would like to ask you the followings; please read them through:

1. Please update your OGB package to the newest one (1.1.1 as of May 5, 2020). We have made a number of changes to the datasets as well as the package. 

2. Please check out our paper (https://arxiv.org/abs/2005.00687) to learn more about OGB.

3. Please keep yourself updated by checking this google group. We will post any major releases or changes here. 

4. Please feel free to post any questions either at here, or at our Github (https://github.com/snap-stanford/ogb) if they are related to the package.

We hereby invite the community to develop and benchmark state-of-the-art graph ML models at https://ogb.stanford.edu.

Best,

OGB Team

[wenjin wang]

Hi, OGB Team,

We are very much interested in your work and want to submit our results to the leaderboard. Could you give us some details about the process?

We would be very grateful if you could provide us with a detailed example.

Best,

Wenjin Wang

在 2020年5月6日星期三 UTC+8上午6:09:32，Open Graph Benchmark写道：

[Weihua Hu]

Hi Wenjin,

Thanks for your interest!

Please see here for the instruction for leaderboard submission. Let us know if you have any questions.

Cheers,

Weihua

[wenjin wang]

Hi Weihua,

Thanks for your reply. 

I have learned the format of the submission. But how could I send these contents to you? Is it ok just to write in the reply on the forum?

Best,

Wenjin Wang

在 2020年5月26日星期二 UTC+8下午6:19:58，Weihua Hu写道：

[Weihua Hu]

Hi Wenjin,

Please see "How to submit?" on the leaderboard page.

You will need to fill in Google form to submit the result.

Best,

Weihua

[wenjin wang]

Hi Weihua,

I have filled out the form.

Thank you very much for your patient guidance.

Best,

Wenjin Wang

在 2020年5月27日星期三 UTC+8上午1:38:23，Weihua Hu写道：

[Weihua Hu]

Hi Wenjin,

Perfect! I have updated the leaderboard.

Thanks for your contribution!

Best,

Weihua

[wenjin wang]

Hi Weihua,

Thanks for your timely attention.

We would be very grateful if you could change the contact name "Wenjin Wang" to "Wenjin Wang (PGL Team)".

Our team will continue to follow up your project.

Best,

Wenjin Wang

在 2020年5月27日星期三 UTC+8下午12:49:25，Weihua Hu写道：

[wenjin wang]

Hi Weihua,

It's nice to see the OGB project getting more attention.

We have a question about the ogbn-proteins dataset. In the Open Graph Benchmark: Datasets for Machine Learning on Graphs, we found that the node feature of proteins is not available. At the same time, there is a specie feature for each node in the dataset. Can this feature be used to train our model? Whether this could lead to a data breach？

Best,

Wenjin Wang

在 2020年5月27日星期三 UTC+8下午12:49:25，Weihua Hu写道：

Hi Wenjin,

[Weihua Hu]

Hi Wenjin,

Great question. Here are the general rules applicable to all of the OGB datasets.

1. Node features are not available for some datasets (which is the case for ogbn-proteins, and also ogbg-ppa etc). 

2. Any features available in the graph object can be used, and there is no data breach.

Best,

Weihua