| description | A practical guide, to creating and understanding network maps |
|---|---|
| cover | ../.gitbook/assets/creative-critical-use-of-networks.jpeg |
| coverY | 0 |
This guide will help you learn how to map and understand networks, equipping you with the tools to visualize complex relationships and interactions within any ecosystem, organization, or knowledge space. Network maps can uncover central actors, organic clusters, indirect relationships, and paths that would otherwise remain hidden. Whether you're a beginner or an experienced mapper, this guide will provide the foundational knowledge to start your journey in network mapping.
Graph Commons stands out from other graph services by emphasizing simplicity and collaborative data exploration in a shared workspace. Here, you and your collaborators can gather data on topics of interest, define and categorize relationships, and transform this data into interactive network maps. Our platform makes it easy to uncover new patterns and share insights on complex issues through a user-friendly interface.
This guide will help you become acquainted with the Graph Commons interface and show you how to create your own graphs in just three steps:
For those interested in a deeper understanding of the value and application of graphs, we recommend reading Creative and Critical Use of Networks on our blog.
Graph is a technical term for network, but also used as “social graph”, “interest graph” etc. A graph on Graph Commons is an interactive network map, and it is always made of two main elements: nodes and edges.
Two nodes connected by an edge
The simplest form of a graph includes just two nodes and one edge. As you add more nodes and edges, the graph expands and becomes more complex.
A larger graph
Graphs can grow indefinitely as more data is added. They can range from small and simple to vast and intricate.
.png)
A nicely sized graph
While it's theoretically possible to create an endlessly expanding graph – akin to the infinite library in Borges's story – practical graphs focus on specific questions or themes.
Next, we'll discuss how to start modeling your graph based on the data and questions you have.
Modeling begins by addressing three key questions:
Identify the entities involved, which can range from individuals and organizations to concepts and objects. An actor is represented as a node, and categorized by node types which are usually singular and broad, like "Person" or "Organization."
Determine how these actors interact. These interactions are the edges between nodes and can be various types like collaborations, transactions, or locations. Examples of edge descriptions include "Person -COLLABORATED WITH- Person", Person -WORKS FOR- Company, or "Company -OWNS- Company." Ensure that edge types are straightforward and reflect the interaction clearly without creating unnecessary complexity.
Choose a title that succinctly describes the scope and purpose of your graph. This helps keep the focus and makes the graph more understandable at a glance. Examples include "NYU Graduates and Their PostDoc Jobs" or "Swedish Changemakers: Impact and Obstacles."
By simplifying your model, you make the graph easier to understand and manage. Once these elements are defined, reflect on your model to see if further simplification is possible without losing essential details.
Before finalizing your model, sketching a network diagram by hand can be very beneficial. Draw nodes (represented as circles), label them, and connect them with lines to represent edges. Use separate colors for different node types and relationship types to enhance clarity and distinction. This manual process helps cement your understanding of the network’s structure. However, as these sketches grow in complexity, they can become difficult to manage. At this stage, transfer your sketch to Graph Commons. Our platform employs a physics-based simulation that automatically organizes the layout based on the connections. This not only showcases the dynamics of your network but also highlights central nodes and peripheral ones, revealing organic clusters that may not have been initially apparent.
Hand-drawn network model of partnerships between French and Turkish NGOs. Image from a Graph Commons workshop.
With a clear model and a defined title, you are now ready to transition from conceptualizing to graphing your data in Graph Commons.
{% content-ref url="creating-a-graph.md" %} creating-a-graph.md {% endcontent-ref %}
After conceptualizing your graph and sketching a preliminary diagram, the next step is to systematically organize your data for import into Graph Commons.
This tool allows you to input and arrange your data directly on the platform. As you populate the table, Graph Commons dynamically creates and updates the visual network map, allowing you to visually adjust nodes and relationships with simple clicks on the map canvas.
{% content-ref url="editing-data-table.md" %} editing-data-table.md {% endcontent-ref %}
For those who prefer to prepare their data in a familiar spreadsheet environment before importing.
- Nodes Table: Each row should list a node with its properties spread across subsequent columns. This setup helps in detailing each node’s attributes clearly.
Each row in the nodes table represents a node's details.
- Edges Table: This sheet should organize the relationships. Each row needs to have the "from" node types and names on the left, the "to" node types and names on the right, with the "edge type" clearly marked in a center column. The column header names are strict. Include columns for additional details like weights if these are part of your data set.
Each row in the edges table represents a relationship.
{% content-ref url="importing-data.md" %} importing-data.md {% endcontent-ref %}
Organizing your data in spreadsheets ensures that every element of your network is accounted for and correctly formatted for smooth transition into the Graph Commons platform. Once your data is structured and ready, importing it into Graph Commons is straightforward, allowing you to see your network take shape in a more sophisticated and organized manner.