In your opinion what is the best approach for link prediction in sparse bipartite networks?
1. Unsupervised approaches ( Such as Similarity Indices )
2. Supervised approaches
3. another ways
If your choice is 1 please point the best method and if your choice is 2 point your best way for feature engineering and if your choice is 3 please explain more.
Seems that it will be a good discussion. ;)