Skip to content

We use gpu to accelerate the influence-maximization diffusion, which can make it faster more than 10 times than before

License

Notifications You must be signed in to change notification settings

SparklyYS/Simultaneous-IMM

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

10 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Compile

use make to compile
"linux & gcc 4.7.2" preferred

Execute

Example:
    ./imm_discrete -dataset nethept/ -k 50 -model IC -epsilon 0.1
    ./imm_continuous -dataset nethept/ -k 50 -model CONT -epsilon 0.1 -T 0.01

Arguments:
    -model:
        the diffusion model uses
        IC or LT or CONT

    -dataset:
        path to the dataset directory

    -epsilon:
        a double value for epsilon

    -k:
        number of selected nodes
	-T:
    	time threshold, only in continous model

Dataset Format

Three files are necessary inside path_to_dataset folder

Format for attribute

Location:
    path_to_dataset/attribute.txt
Format:
    This file should have exactly two lines
    n=number of nodes
    m=number of edges

Format for IC model

Location:
    path_to_dataset/graph_ic.inf
Format:
    Each line has three numbers
    node1 node2 propogation_probability_from_node1_to_node2
Comments:
    It is always a directed graph
    node number should range in [0 to n-1] (inclusive)

Format for LT model

Location:
    path_to_dataset/graph_lt.inf
Format is same as IC model

Format for CONT model

Location:
    path_to_dataset/graph_lt.inf
Format:
    Each line has Four numbers
    node1 node2 param1 param2
For the details of two parameters, refer to http:https://www.cc.gatech.edu/~ndu8/DuSonZhaMan-NIPS-2013.html

About

We use gpu to accelerate the influence-maximization diffusion, which can make it faster more than 10 times than before

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published