GemsLab / REGAL

Representation learning-based graph alignment based on implicit matrix factorization and structural embeddings
https://gemslab.github.io/papers/heimann-2018-regal.pdf
MIT License
92 stars 16 forks source link

Graph alignment problem with different node amounts in two graphs? #8

Open JackieAnxis opened 4 years ago

JackieAnxis commented 4 years ago

Hello, I'm curious about that is REGAL enable to deal with the graph alignment problem where there are different node amounts in two graphs. Thanks!

renzunxiao commented 4 years ago

May I ask if you have ppi and arxiv data sets that can be shared with me? ,my email is rmmluna@163.com

markheimann commented 4 years ago

Hi Jiacheng,

Sorry for the late response. REGAL can deal with this problem—this implementation in fact embeds the graphs jointly by combining them into one large block diagonal adjacency matrix (which you can do with any number and size of graphs). You can also perform the nearest-neighbor search for each node in graph 1 among any number of possible nodes in graph 2. The only thing to change is when we split the embeddings per graph, instead of assuming the first half of the vertices are graph 1 and the second half are graph 2, you’d have to pass in the indices at which to split. Hope this helps!

Best, Mark

On Mar 11, 2020, at 9:23 AM, Jiacheng Pan notifications@github.com wrote:

Hello, I'm curious about that is REGAL enable to deal with the graph alignment problem where there are different node amounts in two graphs. Thanks!

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/GemsLab/REGAL/issues/8, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABCQJQ2ZPE45VB6QJAEEFSLRG6GFLANCNFSM4LFVWBMQ.

JackieAnxis commented 4 years ago
font{
    line-height: 1.6;
}
ul,ol{
    padding-left: 20px;
    list-style-position: inside;
}

Thanks for your response. The process you described works well!Best,Jiacheng

On 03/25/2020 17:29,renzunxiao<notifications@github.com> wrote: 

May I ask if you have ppi and arxiv data sets that can be shared with me? ,my email is rmmluna@163.com

—You are receiving this because you authored the thread.Reply to this email directly, view it on GitHub, or unsubscribe.