Using Gephi: network visualization (tools n’ tips)

cropped-best.png Being able to “Visualize Networks“… Sounds pretty cool and “not even that hard”. But one has to know what tools ot use and how to use them. Here are a few tips on how to use Gephi (a network visualization software).

If you are in the Data Viz’ Biz’ -as a journalist or a researcher-, you already know about visualization tools. It is the (not so) new trend, the thing cool kids on the block do all the time. Basically they help you understand an issue or a current event by creating a representation of it. Even TheGuardian tells you how to create your own data visualisations. Most journalists and researchers use mapping or graph tools and softwares, I personally have a preference for network visualisations, which I define as an abstract way to look at the many connexions between things/people/ideas/texts/… in a qualitative or quantitative manner.

I have been using Gephi for mainly four reasons: it runs on Windows AND Mac, it’s free, open-source, and it’s made by other researchers. Gephi is a network visualization software that allows you to work on networks -may they be social or not and digital or not.

DISCLAIMER: this article is not a tutorial (persee) on how to use the software, if that’s what you’re looking for, just go there.

You can either use an automatically generated database (by example using a Web crawler, see point next section) or build one yourself. It as easy as creating an excel spreadsheet with to columns such as:

Col1  A   B   B   B

Col2 B   C   D   E

In this example, B can be regarded as a node connecting entities A, C, D and E (see pictures below). What I propose today is to give you some tips on how to use Gephi -things that I sometimes have learned the hard way-, as well as some tools that go around it.


1. Save regularly

This might sound obvious to regular Gephi users, but if you just started to use the software, know that there is no back button. If you generate the wrong spacialization or move many nodes manually, you won’t be able to go back. What I usually do is save the file before doing anything that I don’t know about. Don’t worry, you’ll get use to it quick. Saving regularly will also enable to go back to previous spacializations to see directly which one shows things in the clearest manner or to compare improvements/changes to your network.

2. You can fill data in yourself…

As I explain earlier, you can fill data in yourself in the data laboratory part of the software (see below). This is a pretty interesting feature because it allows you to create representations of non-digital, non-social networks. It could be interesting to, for example, generate a representation of the people working in your company -in a cool slick way. Capture d’écran 2015-01-20 à 18.58.25  Capture d’écran 2015-01-20 à 18.58.37       Capture d’écran 2015-01-20 à 19.00.09

3. (including TimeStamps)

This is another cool feature of the software. You can include Timestamps within you data to study the evolution of your network over time. Sadly Gephi does not allow you to export it as video (yet) but you can use another software to record the video, or make multiple screenshots and create a .GIF file like this: http___makeagif.com__media_12-18-2014_ZhU_I8

4. … or insert some!

I can’ stress that enough: Gephi let you do whatever you want with your data. That means that you can use data you collected on social media and then add nodes and links of your own you deem it relevant. This allows researchers (or journalists) to do in-depth, multi-level analysis.


1. Combining Gephi and a Web crawler for for links analysis

This is my favorite combinaison so far. By combining a Webcrawler such as the “NaviCrawler” of WebAtlas, you become able to create visualizations of websites and the links between them. This means that you can how websites are connected to each other, which are linked and vice versa. Here is a tutorial (in french) on how to do so. Once you get used to it, the process becomes quite simple, and can be repeated infinitely:

Capture d’écran 2013-12-29 à 20.12.37 Capture d’écran 2013-12-29 à 20.12.55

2. Combining Gephi and this app to visualize your social network

If you want to have visualization of your social network (in this case Facebook), you can use this app. Granted this is a pretty self-centered thing to do, but you can use similar principales to map let’s a the evolution of a particular topic on twitter or other social-networking sites. Here is my social network, it is pretty interesting that the different nodes (or area where many entities are concentrated) represent my different groups of friends (from high school, the university, my travels, my family, etc.). 10714276_10152319616535728_631975038985409003_o

3. Combining Gephi and Inkscape to make your graphs sexy

Inkscape is a pretty cool software. I’ve only started to use it, but I guess it give pretty good results. Think of Inkscape as a Photoshop for your visualisations. Once you have exported your Gephi networks, you can use Inkscape to change the colors and size of nodes, links and much more.

One last tip: As my teachers (and now colleagues) always tell me, never forget that the tools you use must stay what they are : tools. If it doesn’t serve a particular purpose, there is probably no use for them. I’m pretty sure this tip works for journalists as well as researchers.   I’ll try to discuss other tools such google fusion tables or maps and other mapping softwares soon. Let me know if you have any ideas! Also, don’t hesitate to tell about your uses of Gephi and other visualization softwares. Other Tips and Tools welcomed!!!

Ps: If you speak French, and want to see some really cool stuff, go there:

2 thoughts on “Using Gephi: network visualization (tools n’ tips)

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s