Graph

buildGraph()

Constructs a graph that has all the automatically recoverable KGs as nodes and the respective relationships with the other KGs as edge.

Returns:

Name	Type	Description
Graph		A graph with all the KGs automatically discoverable with its relationship.

Source code in kgheartbeat\Graph.py

def buildGraph():
    """Constructs a graph that has all the automatically recoverable KGs as nodes and the respective relationships with the other KGs as edge.

    Returns:
        Graph: A graph with all the KGs automatically discoverable with its relationship.
    """
    allKg = AGAPI.getAllKg()
    idList = []
    G = nx.Graph()
    for i in range(len(allKg)):
        element = allKg[i]
        id = element.get('id')
        idList.append(id)
    for j in range(len(idList)):
        externalLinks = aggregator.getExternalLinks(idList[j])
        exLinksObj = utils.toObjectExternalLinks(externalLinks)
        print(idList[j])
        if isinstance(exLinksObj,list) and len(exLinksObj) > 0:
            for k in range(len(exLinksObj)):
                link = exLinksObj[k]
                value = str(link.value)
                value = re.sub("[^\d\.]", "",value)
                if value == '':
                    value = 0
                value = int(value)
                G.add_edge(idList[j],link.nameKG,weight=value)
    #pos = nx.spring_layout(G, k=0.8)
    #nx.draw(G,pos,with_labels=True,width=0.4,node_size=400)
    #plt.show()
    return G

getCentrality(graph, idKg)

Compute the centrality of the node in the Graph of all KGs.

Parameters:

Name	Type	Description	Default
graph	Graph	A Graph that represent A graph with all the KGs automatically discoverable with its relationship.	required
idKG	string	A string that represent the id of the KG whose page rank we want to calculate.	required

Returns:

Name	Type	Description
float		A float that is the centality of the KG in the graph og all KGs.

Source code in kgheartbeat\Graph.py

def getCentrality(graph,idKg):
    """Compute the centrality of the node in the Graph of all KGs.

    Args:
        graph (Graph): A Graph that represent A graph with all the KGs automatically discoverable with its relationship.
        idKG (string): A string that represent the id of the KG whose page rank we want to calculate.

    Returns:
        float: A float that is the centality of the KG in the graph og all KGs.
    """
    degreeCentrality = nx.degree_centrality(graph)
    return degreeCentrality.get(idKg)

getClusteringCoefficient(graph, idKG)

Compute the clustering coefficient of a KG in the graph of KGs (the higher it is, the more it means that the node's ability to form a clique is high).

Parameters:

Name	Type	Description	Default
graph	Graph	A Graph that represent A graph with all the KGs automatically discoverable with its relationship.	required
idKG	string	A string that represent the id of the KG whose page rank we want to calculate.	required

Returns:

Name	Type	Description
float		A float that is the clustering coefficient of a KG in the graph of all KGs.

Source code in kgheartbeat\Graph.py

def getClusteringCoefficient(graph,idKG):
    """Compute the clustering coefficient of a KG in the graph of KGs (the higher it is, the more it means that the node's ability to form a clique is high).

    Args:
        graph (Graph): A Graph that represent A graph with all the KGs automatically discoverable with its relationship.
        idKG (string): A string that represent the id of the KG whose page rank we want to calculate.

    Returns:
        float: A float that is the clustering coefficient of a KG in the graph of all KGs.
    """
    clusteringCoefficient = nx.clustering(graph,idKG)
    return clusteringCoefficient

getDegreeOfConnection(graph, idKg)

Compute the degree of connection of a KG (number of edges).

Parameters:

Name	Type	Description	Default
graph	Graph	A Graph that represent A graph with all the KGs automatically discoverable with its relationship.	required
idKG	string	A string that represent the id of the KG whose page rank we want to calculate.	required

Returns:

Name	Type	Description
int		A integer that is the number of edges of the KG.

Source code in kgheartbeat\Graph.py

def getDegreeOfConnection(graph,idKg):
    """Compute the degree of connection of a KG (number of edges).

    Args:
        graph (Graph): A Graph that represent A graph with all the KGs automatically discoverable with its relationship.
        idKG (string): A string that represent the id of the KG whose page rank we want to calculate.

    Returns:
        int: A integer that is the number of edges of the KG.
    """
    degree = graph.degree(nbunch=idKg)
    return degree

getPageRank(graph, idKg)

Compute the PageRank of a KG. This is a value in range [0;10].

Parameters:

Name	Type	Description	Default
graph	Graph	A Graph that represent A graph with all the KGs automatically discoverable with its relationship.	required
idKG	string	A string that represent the id of the KG whose page rank we want to calculate.	required

Returns:

Name	Type	Description
float		A float that is the PageRank of the KG.

Source code in kgheartbeat\Graph.py

def getPageRank(graph,idKg):
    """Compute the PageRank of a KG. This is a value in range [0;10].

    Args:
        graph (Graph): A Graph that represent A graph with all the KGs automatically discoverable with its relationship.
        idKG (string): A string that represent the id of the KG whose page rank we want to calculate.

    Returns:
        float: A float that is the PageRank of the KG.
    """
    pr = nx.pagerank(graph)
    return pr.get(idKg)

storeEdges(graph, nodelist)

This function store info about all edges for every KG in the graph with all the KGs automatically discoverable (this is used for visualize the graph in the web app).

Parameters:

Name	Type	Description	Default
graph	Graph	A Graph that represent A graph with all the KGs automatically discoverable with its relationship.	required
nodelist	list	A list that contains all the IDs of the KGs whose subgraph we want to extract.	required

Source code in kgheartbeat\Graph.py

def storeEdges(graph,nodelist):
    """This function store info about all edges for every KG in the graph with all the KGs automatically discoverable (this is used for visualize the graph in the web app).

    Args:
        graph (Graph): A Graph that represent A graph with all the KGs automatically discoverable with its relationship.
        nodelist (list): A list that contains all the IDs of the KGs whose subgraph we want to extract. 

    """
    save_path = './Graphs Visualization JS/Subgraphs'
    for i in range(len(nodelist)):
        newFilename = re.sub(r'[\\/*?:"<>|]',"",nodelist[i])
        remove_punctuation_map = dict((ord(char), None) for char in '\/*?:"<>|')
        newFilename = newFilename.translate(remove_punctuation_map)
        remove_punctuation_map = dict((ord(char), None) for char in string.punctuation)
        newFilename = newFilename.translate(remove_punctuation_map)
        completeName = os.path.join(save_path, newFilename+".txt")
        e = graph.edges(nodelist[i])
        e = str(e)
        e = e.replace('(','[')
        e = e.replace(')',']')
        e = e.replace("'",'"')
        with open(completeName,'w',encoding="utf-8") as f:
            f.write(e)