About Graph Commons
Graph Commons is a collaborative 'network mapping' platform to create, share, navigate, and learn about relationships among people, organizations, or concepts. People create any type of connection between the things they relate to, publicly or privately.
With Graph Commons, you can:
* Create network maps with friends or experts to get insights about a topic
* Connect your data to others' to see a bigger picture
* Browse graphs that you are in or related to your interests
* Follow topics, graphs, and people to create your personalized feed
Graph Commons is being developed by Burak Arıkan and Cenk Dölek.
What connects? is the question to start modeling a network. While nodes can represent a person, an organization, or a concept, connections fall into the following categories:
Transmission Networks – Something actually flows. Water flows, electricity flows, money flows, news flows... Usually physical, and it could be broken like a pipe.
Interaction Networks – Connection is an event, with a specific time. I email you, I buy something, we do an exhibition together... Something passed during a contact. Explicit.
Attribution Networks – Connection is an expression of a relationship. You are my friend, I love you, you trust him, she recognizes you... Visible only if you state it.
Affiliation Networks – Connection is a belonging to a group or category. We are in the same school, things are in the same category, organizations connected by board members... Linked by correlation, similarity, or membership. Implicit.
Where to start a graph? is a common question when making network maps. After you imagine some relationships, you can start with the first thing that comes to your mind, then grow and tweak the map step by step from there.
Where to stop a graph? is another usual question that comes quickly, especially if you feel like everything is connected to everything else. Putting a definitive graph title and considering only the strong connections help to limit your network map's scope.
Connection weight affects the network formation dramatically. Weight of a connection could be frequency of interaction, number of items exchanged, individual perceptions of strength of the relationship, costs in communication or exchange, or combination of these. Strong connections bring closer the two end nodes, and reveal tight clusters. In fact, strong ties are more transitive than weak ties.
Collaborative mapping enables more fruitful and complete graphs, in fact, it is great for brain storming. Graph Commons enables collaborative graph editing. As you flow in a visual discussion, a network is being mapped.
After a network is mapped, you can learn from it in various ways. To understand networks, we measure the centrality of actors. It gives us insight into the various roles and groupings in a network -- who are the connectors, leaders, bridges, isolates, where are the clusters and who is in them, who is in the core of the network, and who is on the periphery? With network analysis and visualization we can navigate the complexity and develop insight through its solid metrics and dynamics.
Centrality – Who are the most important actors and what are their locations in the network? You can prioritize actors based on their position and connectivity in the network. Centrality measure is not just the number of connections an actor has. How much a bridge role it has between clusters, and how close it is to other members of the network.
Clusters – What organic groups or clusters exist in a network? Discovering clusters of nodes, which have more connections to one another than they do to outsiders. You can name the clusters by looking at who is in it. The structural holes between the clusters as well as the bridges between them are as important as the clusters themselves.
Equivalency – Which actors are alike? Determining actors who play a similar role and have similar positions. They show us alternative paths, as well as redundancy in the network. Refers to the extent to which actors have a common set of linkages to other actors in the system.
Shortest Paths – What is the distance between two actors? What indirect relationships exist? Revealing normally invisible connections and the degrees of seperation between actors.
Density – How well a network is connected, compared to other networks? Comparing density of networks, as well as connectivity of different regions within a single network.
Diameter – What is the longest path in the network? Finding the reach, that is, how long it will take at most to reach any node in the network.