Social Network Analysis

Usage

You can create conda env from file for convenience: conda env create -f conda_env.yml.

data/: data files should be there as names provided from zip. Full data files are in .gitignore, only samples in git for quick testing stuff. May add more samples
src/
- notebook/: notebook stuff, presenting etc
- pyhton/: add modules there

do Todos (R DONE)
implement modularity maximization (R DONE)
Strache bug fix + deep dive (V DONE)
Community von einem Artikel -> Nodes davon nehmen -> Lookup auf Following Graph -> Similarity berechnen (O DONE)
Metriken finden um zwei Communities zu vergleichen (O, G DONE)
Plot similarity between timestamps and "final state" = last timestep (scale 0-1) (V DONE)
- needs Similarity measure = #interactions (similar to line plot in semantic.ipynb)
Line Chart: plot relative interactions for each community over time (#interactions / #members), similar to lineplot in semantic.ipynb (G DONE kinda)
Cyclic behavior -> do above plot each day and plot them on top of each other (R DONE)
- add vertical lines for biggest articles (R DONE)
Präsi (G)
Report
- theoretical background
  - community detection (mod maxim) (R)
  - anderer der langsamer war: Girvan-Newman (O)