Applications of networks analysis: from mapping entrepreneurship endeavours to visualising your LinkedIn connections

Daniela Valenzuela — Innovation Consultant

As part of my MSc in Strategic Innovation Management, I took the Network Analysis and Infographics module. I have always known about the importance of building your network and networking; however, I never really seized the immense value that analysing a particular network brings.

What is network analysis?
In its simplest definition, a network is formed by nodes, the entities that are analysed and links, the ties between the entities.

Mapping the relationships between the entities can be achieved through network analysis; when we look at individuals, groups and communities, we are talking about social network analysis. We rely on combined disciplines for this analysis drawing from mathematics, anthropology and psychology. From mathematics, we use graph theory to describe and represent networks; this allows us to use measures, which can be essential to compare networks and extract powerful insights. However, numbers fall short, and we need to rely on the social sciences to interpret relationships; anthropology helps us understand the behaviours between human beings, and psychology uses sociometry and explores how our minds work.

All of this makes sense, but I would say this falls into place when you are able to visually represent a network, which I find incredibly pleasant, given how complex and unpredictable social relationships are. When representing a network, you can start to see the actors in your focus of analysis and, of course, their connections.

Figure 1. Example of a simple directed network with adjusted node size and colour according to categories.

With this, you can identify network level measures such as the connectivity of a network (density), the presence of unconnected nodes (isolates), the presence of unconnected groups (components), highly connected sub-groups (cliques), a measure of the average resources needed to reach all nodes (average path length) and many more. All of these measures are most insightful when compared with other networks. Now, within a network, you can analyse node level measures such as the most connected entities (highest degree), how fast a node can reach all others in the network (closeness) and how likely a node is part of the most direct route between nodes (betweenness).

These measures, when interpreted can provide information about popularity, information flow, influence, speed, efficiency, control and brokerage. A holistic view indicates a network’s dynamics, diffusion and resilience.

Applications of network analysis

Flights
The applications of network analysis are multiple, for example, looking at the flight connections between airports. The nodes are the airports and the ties are the flights which are represented in a layover of a geographical map. This gives us a quick snapshot of the areal movement of people across the world. We can see hubs of connection in places like Europe, the US and East Asia.

Figure 2. Connections between 3200 airports [http://www.martingrandjean.ch/wp-content/uploads/2016/05/airports-map.png]

Crime
Another application which I find fascinating is looking at networks of crime, for example, drug dealing groups. Police intelligence uses these types of techniques to identify groups and understand where to attack the network to break it or damage it as much as possible so the flux of drugs can stop. In network lingo, this is called bridging opportunities.

Figure 3.CIA Map of International illegal drug connections.

Innovation
Now, what is the importance of network analysis for innovation?

Because innovation is the process of creating and delivering new value, combining resources, knowledge, and skills is essential; this means innovation is inherently interdisciplinary and multi-functional. We see that the most amazing inventions[1] that have reached the public come from a combination of backgrounds and knowledge.

In static environments, innovation does not flourish, whereas, in dynamic ones, we see the prosperity of new products and services, which often are embedded in a rich ecosystem of start-ups, large corporations, incubators, accelerators and venture capital. When we are looking to purposely incentivise innovation, public and private initiatives look at addressing ‘network failures’ by enriching innovation ecosystems, which in other words, means creating or nurturing a network. It is not enough to have links; they need to be productively aligned.

By mapping the interactions in a particular innovation ecosystem, we can develop a level of understanding of the connectivity of the network as well as identify disconnected actors. Even more so, when we map a network over time, we can see the evolution of interactions, which usually can be a good indicator of knowledge transfer over time between actors.

One example which allows us to look at the relations of organisations collaborating under UKRI grants is in figure 4. This essentially displays the relationships of collaboration in innovation projects.

Figure 4. Network of projects under UKRI grants. Developed by the author.

Here we can observe the relationship between organisations collaborating under publicly funded UK Research and Innovation (UKRI) grants. The size of the node was adjusted relative to the quantity of funding received. Here, we can analyse several elements, but at a glance, we can see universities and hospitals collaborating together; we also see a scarcity of private sector representation and isolated organisations working disconnected from others. A visualisation like this can tell us so much about relationships, which could even inform changes to the way financial instruments are structured, for example, by generating incentives for the private sector to take part.

Entrepreneurship Endeavours

A novel application developed by this author, in collaboration with my MSc classmates, is to analyse a network before intervening in it in the form of entrepreneurship. It would be incorrect to think that anyone starting a venture being a social or profit initiative is alone in an ecosystem; there will always be an existing network of actors. When looking to make a contribution seeing the pre-existent actors and their relationships is key to understanding how to fit in — or disrupt- to achieve real impact. In this case, we were looking at understanding the network of woodland creation initiatives since we wanted to develop a tree planting service in Sussex.

We quickly found the most connected actors, highlighted with the labels, connected to several smaller organisations which are actually plating the trees. We identified the key clusters, as well as those organisations working in isolation.

We mapped the flow of money that promotes tree-planting endeavours. We learned that there was an active network, and there was indeed space for another partner to facilitate the process, which confirmed our willingness to take part in the network and contribute to more connections.

Figure 5. Network of Woodland Creation Initiatives around Sussex. Developed by the author.

LinkedIn Mapping

A last but obvious application is looking at our connections on LinkedIn. Thanh To developed an easy demo to plot your LinkedIn connections. First, you download the dataset directly from LinkedIn, and then you upload it to Thanh’s demo. Here you see a layout of my connections; I have replaced the names with codes for personal purposes. Those working in the same organisation are connected and have the same colour; at the centre in blue, there is me holding the ego network together. From this, I learned that my alma mater is an essential part of my network, highlighted in the big cluster in green.

All those people are working at my bachelor’s University currently and are densely connected. This could be the very start of understanding your professional networks.

Figure 6. LinkedIn connections based on author connections using https://devpost.com/software/linkedin-connections-visualization

I guess we all map networks in our heads; we know who are the most important actors in our field, and we know who works with who to a certain extent. However, its visual representation provides such a clear picture, which I believe could be a game changer when looking to create endeavours and initiatives that promote systemic change.

If you want to talk about using Network Analysis for innovation purposes, please reach me!

[1] Innovation is not the same as invention. Innovation is the implementation that creates value.

References:

Main content based on:
Rotolo, D. (2022). 959N1: Network Analysis and Infographics.

LinkedIn Connections Visualisation: https://devpost.com/software/linkedin-connections-visualization

Special thanks to Anas Aleassa for having rich conversations about network analysis with me.