DEV Community

hub
hub

Posted on

diving into Data-Analysis with Gephi - some first steps...

diving into Data-Analysis with Gephi

here some steps:

diving into Data-Analysis with Gephi - some first steps...-
feel free to add your ideas now..

1. Collecting Data:

a. Twitter API:
Consider using the Twitter API to gather more comprehensive data on followers and followings. This will allow you to collect more data than what you can manually obtain from the web interface. The Twitter API will require setting up a developer account, creating an application, and obtaining API keys.

b. Manual Collection:
If you prefer a simpler approach, manually copy-pasting data from Twitter profiles can serve as a starting point. However, keep in mind the limitations in terms of the amount of data you can collect.

2. Data Preparation:

a. Format Data:
Organize your data into a format suitable for Gephi. You may want to create a CSV file with columns like 'Source' and 'Target' representing follower-following relationships.
b. Additional Data:
Consider adding additional columns for attributes (e.g., user details, account creation date) that could enhance your analysis.

3. Import Data into Gephi:

a. Open Gephi:
Download and install Gephi if you haven't already. Open the application.
b. Import Data:
Use the 'Data Laboratory' tab to import your CSV file. Assign columns to 'Source' and 'Target.'

4. Explore Gephi Features:
a. Layout:
Experiment with different layout algorithms to arrange nodes in a visually appealing way. ForceAtlas2 is a popular choice.
b. Node Size and Color:
Adjust node size and color based on node attributes. For example, you could color nodes based on account creation date.
c. Filters:
Utilize filters to focus on specific parts of the network, such as influential nodes or communities.

d. Statistics:
Gephi provides various statistics to analyze the structure of the network. Use these to gain insights into network characteristics.
e. Preview:
Use the 'Preview' tab to visualize the network. Adjust settings for a clear representation.

5. Iterate and Explore:
Add more Twitter accounts to your dataset and repeat the process. You can then compare and contrast different networks or combine them for a more comprehensive analysis.

  1. Documentation and Resources: Refer to Gephi's documentation and tutorials for more in-depth knowledge about its features: https://gephi.org/users/ Remember to respect Twitter's terms of service and privacy policies while collecting and analyzing data. Happy exploring!

*here a concrete example. *

-----Dive into the work with GEPHI - via social network analysis methodology, i want to work on a (very simple) dataset to get some first visualization and analysis

whats the best way to headstart into the State of the art, and workflows, and best practices:

i have heard about gephi: well can i start with a very simple dataset - eg. the twitter dataset

for example: we take *this dataset: *

a. https://twitter.com/hwk_konstanz

and if we take here the followers:

b.https://twitter.com/hwk_konstanz/following

what about the visualizaition of the followings (or followers) of the above entities

@Holzindustrie
Follows you
Der HDH ist der Hauptverband der Deutschen Holzindustrie und Kunststoffe verarbeitenden Industrie und verwandter Industrie- und Wirtschaftszweige e.V.
vbw-bayern

@vbw_bayern
Nachrichten der Bayerischen Wirtschaft Impressum: http://vbw-bayern.de/vbw/Impressum/index.jsp
BWA Deutschland

@BWA_Vorstand
Offizieller Account des Bundesverbandes für Wirtschaftsförderung und Außenwirtschaft (BWA)
Bayernwerk

@Bayernwerk
Die Bayernwerk-Gruppe gestaltet gemeinsam mit den Menschen in Bayern die Energiezukunft im Freistaat aktiv mit. Impressum: http://bayernwerk.de/impressum
MittelstandsWiki

@mittelstandnews
Themen für Unternehmen. Impressum: https://miwiki.de/imprint
Deutschlandfunk

well if i start to visualize the followers (or the followings) of this account, we can just copy-paste them in a text or table editor.
It will give us a very simple "flower", a network with our account in the center and all the followers around it, not very interesting
but if we do the same with a few other accounts we'll get overlapping "flowers", a network that will be like a Venn diagram.

  • and will play with the data - first of all i will add some data - the

a. https://twitter.com/hwk_konstanz/following
b. another twitter-account- /following
c. another twitter-account- /following
d. another twitter-account- /following
e. another twitter-account- /following
f. another twitter-account- /following

any further ideas to start with Gephi using twitter-data!?
... many thanks in Advance - i will the first steps ..

look forward to hear from you

regards

Top comments (0)