/experience-oim-under-dcm

experience for paper online influence maximization under decreasing cascade model

Primary LanguagePython

experience-oim-under-dcm

experiment environment

python 3

dataset

  • put raw file flickrEdges.txt and netHEPTEdges.txt under direction \DataProcess\raw\
  • run DataProcess/SampleSmallGraph.py , SampleSubGraph.py to generate network
  • run files under direction SampleFeature to generate related features

main

under the root directory, run

python3 Syn.py

the results will be saved under \SimulationResults\[dataset]\[alg], which can be declared in Syn.py

the same for Fixed.py and Interval.py

draw results

under the visualizationTools, run

python drawResults.py

you can specify:

  • fileFolderPath: where the results you want to draw stored
  • alg_list: the baseline algorithms you want to draw, in the form: ([dir_name], [colum_tag], [coclor], [label])
  • count: the rounds you want to draw
  • issave: save the file or not
  • file_name: the name of your stored file
  • subTitle: the title of your picture

file structure

. 
| BanditAlg # Baselines for running influence maximization problems.
| DataProcess
   ├── raw # store raw file describe the graph
   ├── SampleSmallGraph.py # sample syn graph
   ├── SampleSubGraph.py # sample sub graph from large dataset Flickr/NetHEPT
| datasets # store each dataset's graph and useful feature
| Model
   ├── IC.py, DC.py # run on IC/DC model to get reward
| Oracle
   ├── CMAB.py, Greedy.py, Greedy_IC.py # offline oracle for CMAB, DC and IC
| RunTools # function to run on different baseline
| SampleFeature
   ├── FeatureVector.py # 
   ├── Indegree.py # count the indegree of every node in graph
   ├── NodeFeature.py # generate
   ├── Probability.py # generate interval or fixed probablity
| SimulationResults # store the results of experiment
| Tool # helpful tools
| visualizationTools # draw the result, count the average
| Syn.py, Fixed.py, Interval.py # call RunTools to run choosen baseline