Input network:

The first character of each node represents the category of the node, including 'a', 'p' and 'c' in our method.
The input edge must be: node blank node.
The input edge must be two class: four node relationships(p-a, p-c, a-c and a-a) and two node relationships(p-a and a-a).
The obtained seed set contains some influencial nodes with 'a' type.
Besides MAHE-IM, other network embedding methods are provided to calculate influence maximization, including DeepWalk-IM, node2vec-IM, LINE-IM and SDNE-IM.
If your network is homogeneous, you can set all the nodes as type 'a' and use the DeepWalk-IM method or node2vec-IM method or LINE-IM method or SDNE-IM method.

Or upload txt file here:


Select method:


If you select MAHE-IM method, please set weight β and relevancy threshold V:

Job ID