Hello, any recommendation on the network embedding works for bi-partite (weighted) networks.
most existing network embedding cannot handle bi-partite graph for recommendation very well, maybe you can have a check on "BiNE: Bipartite Network Embedding"
we usually use network embedding to build i2i similarity, and then to use CF method to generate the recommendations like: item2vec, entity2rec
but, in my opinion, factorization-based solution can be deemed as a type of network embedding
oh, "Asymmetric Proximity Preserving graph embedding" is another choice of network embedding for making recommendation
Can we start a discussion on the potential reorganization, table of contents and so on. I believe that there are huge potential gains regarding the reorganization. In the last few days I linked the repository in a number of other awesome aggregator lists.
I'm working a related project and will handle this later. I'll get back to you afterwards. It's still welcome if you'd like to share your idea of how to compile the list in a better way.