Repository of Fair Group Summarization with Graph Patterns

Full Version: https://github.com/PanCakeMan/FGS/blob/main/full.pdf

Dataset

Covid-19: https://github.com/imdevskp/covid-19-india-data/

Movie: https://www2.helsinki.fi/en/researchgroups/unified-database-management-systems-udbm
DBpedia:https://www.dbpedia.org/resources/knowledge-graphs/

• To identify node types, we integrate DBpedia Ontology for building movie data graph. Please see: http://wikidata.dbpedia.org/services-resources/ontology

Linkedin: Cosnet: Connecting heterogeneous social networks with local and global consist

Test Case example (DBpedia Movie):

Call Procedure: VF2Checker.runFGS or VF2Checker.runFGSStream.

Test cases are commented in the main function.

You can commment out to investigate the details.

Example:

1. Build graph (para 1: ontology of DBpedia file , para 2: Movie graph file, para 3: IMDB enrich data file ):

    MovieDataGraph graph = new MovieDataGraph("newType.ttl",
             "mix.dbpedia.graph", "film.imdb.json");
   

2. Define constraints (two movie groups with range constraint [20,30]):

    ArrayList<String> group = new ArrayList<>();
    group.add("Comedy");
    group.add("Action");
   
    ArrayList<Rangepair> cc = new ArrayList<>();
    cc.add(new Rangepair(20, 30));
    cc.add(new Rangepair(20, 30));
   

3. Call proecedure (para 1: group (group), para 2: constraints(cc), para 3: hop(2), para 4: number of node in Vp(m=50), para 5: size of result pattern set (n=10))

   checker.runFGS(group, cc, 2, 50,10);