What is more important than data consolidation? Making sense of that data and communicating it to others. With data volumes growing and time for analysis decreasing, this doesn't seem like an easy task. Data visualisation, however, may be the most reliable way to get on with this daunting process and make it seamless. Data visualisation and more specifically graph visualisation helps you organize all your data and make it readable and insightful.
What is graph visualisation?
Before explaining the essence of graph visualisation we must first define what is a graph.
It consists of nodes and edges where the nodes represent a specific data point, and the edges show the connections between the nodes. For example, nodes might represent a person, a company, a movement, etc. and the edges will show how each of these relate to each other (as in person X is working in company Y).
In addition, further properties of information can be captured and visualised for both nodes and connections. For instance, the gender of individuals could be represented with varying symbols or color schemes.
Ultimately, every researcher, whether in the intelligence field or otherwise, seeks to provide actionable insights to their audience.
But finding and communicating those insights through long and blocky walls of text don’t appeal to decision-makers that rarely have the time to read the report beyond the executive summary.
As a result, information has long been organised in tabular format. Such structure of the data and information has enabled organisations to create all sorts of data visualisation to support decision-making (pie charts, trends charts…). However, such an approach to data storage leads to information silos. This segmentation of the information in different storage areas means organisations deprive themselves of critical insights.
Graph visualisation, on the other hand, makes use of the human brain’s ability to recognise visual patterns easily. It translates large volumes of data into simple images to provide you with a quick and clear understanding of it.
This in turn helps you identify hidden insights and draw better conclusions ultimately saving time and energy.
What’s more, graph visualisations are often dynamic, allowing their users to manipulate data in a way convenient for them. Adjusting colors, shapes, depth of queries and relationship types allows for the creation of a tailored environment which means that every researcher can personalise their graph as they see fit.
When done right, visualisation acts as a bridge between connected data and analysts allowing for a deeper understanding of a problem or a situation and empower decision-makers to respond to the insights in a timely and appropriate manner.
As Tufte describes, graphs should, among other things, “Induce the viewer to think about the substance, rather than about methodology, graphic design, the technology of graphic productions, or something else.” Graphs should “Present many numbers in a small space, make large data sets coherent, and encourage the eye to compare different pieces of data.”
What are the benefits of graph visualisation?
- 1. Trends and relations identification
Humans are more likely to detect important information when it is presented visually rather than textually. Emerging trends, invaluable connections, patterns, and specific data points may easily get lost in the ocean of data that’s surrounding any thorough analysis.
An increasing number of decisions are based on insights derived from connected data. And one effective way to present this data is in graphs. They are easy to construct, seamless to work with, and grant the user with a better and clearer understanding of the data displayed.
In particular, analysing the evolution of a particular network overtime can yield powerful insights. For instance, understanding how the business connections of a given company and the personal connections of its leadership evolve over time can tell us a lot about the potential for fraudulent activities or risk to its reputation.
2. Analysis on various level of detail
Visualising your data through dynamic graphs means that you can control what’s shown and what’s not. Also, you can manipulate what you see by changing the layout, forming different clusters of nodes, or simply showing connections up to a certain depth of relation. This allows for a controlled multi-layered analysis of data at a level needed by the user and according to their goal.
3. Flexible delivery of results
Graphs allow the user to insert additional data or properties into an existing graph without complications or loss of functionality as it is sometimes the case with tables.
What’s more, graphs enable storytelling and better delivery of results. While insight discovery is the foundation for making a modification, the decision-maker is still the ignition that will start the engine of change. Therefore, even if the analyst had done their job extremely well analysis-wise if they are unable to convey what’s important to the one in charge of the decisions, all previous work is not that meaningful.
Graph visualisation, on the other hand, provides the analyst with the tool to simplify insight delivery and communicate complex ideas in a simple manner.
4. Timely and accurate decision-making
This better communication reduces misunderstandings making the whole decision-making process smooth and efficient. When information is easily accessible and readable in the form of graphs, those in charge can quickly find support for their decisions and undertake timely actions. Timely actions, in turn, result in a more efficient and effective workflow providing a harmonious working environment for everybody involved in the process.
5. Storytelling and shareability
Storytelling is the best way to convey a message and make it memorable. If the findings of a piece of research need to be communicated to a wider audience with diverse interests and knowledge on the matter, graph visualisations can surely be utilized.
In fact, great visualisations can deliver great stories. If a graph is well-made it is likely to communicate its key findings in an engaging and coherent manner, thus, reaching more people and having a greater impact.
Advances in graph databases have made graph visualisation increasingly mainstream. And a growing number of companies rely on connected data to perform their analysis and act on their insights.
Such a method is increasingly used for intelligence, due diligence, anti-money laundering, financial fraud detection, and investigative journalism because it is great at merging multiple dimensions while dealing with discrete pieces of evidence.
Finally, by using the power of visualisation, such methods allow researchers to find insights faster and communicate those to decision-makers in a compelling way.
Graph visualisation vs. Spreadsheets
I picked up a few characters from the well-known Spanish series Money Heist and visualised some information about them. I included their role in the movie, their real names and age, and most importantly, their interpersonal relationships during the first season. For the purpose of this article, I did that through both a table and a graph and compared them.
The first spreadsheet of mine looked somewhat inarticulate and had loads of duplicate cells.
I figured out that to make it more coherent I may need to divide the information into two tables.
The first table displayed the characters' basic info clearly. However, the second one turned out ambiguous.
I simply wanted to display how five of the main characters are related to each other and what are the types of their relationships.
As previously mentioned, graph visualisation offers a better way to present connected data in comparison to spreadsheets. That's why I created a graph using the same set of information.
Not only did I get a clearer image of the relationships between the fictional figures but I could also further adjust my graph to reflect specific connections.
I customised my graph to emphasize who is the leader and who are his direct subjects. And if I want to know more about The Professor, his node reveals everything else upon a click.
All graph visualisations were made using reKnowledge. If you want to learn more about it, click here.