About this Talk
Network science has rapidly expanded over the past two decades as a powerful subfield of data science, placed at the intersection of mathematics, physics, social sciences, and computer science. Today, it allows practitioners to extract, analyze, and even predict the network structures underlying virtually any complex system—from social networks and finance to urban planning and healthcare.
In this talk, I will showcase the versatility of network science through practical examples, illustrating its broad relevance and real-world applications.
Description
Network Visualization in Practice with Gephi
Network science has become one of the standout disciplines of the past two decades, with its applications ranging from recommendation systems to HR, from knowledge graphs to social networks.
Visualization is one of the primary components of quickly observing, analyzing, and interpreting complex networks and graphs, and it usually provides a wide variety of insights for future analytics, such as predictive modeling and cluster analysis.
Hence, the main goal of this masterclass is to get participants on board with network visualization using the widely popular open-source software, Gephi. Learners will be provided with a previously prepared graph object, which will show the similarity graph of the conference participants derived from our online survey. In this network, each node corresponds to a survey respondent, and two nodes are linked if those two - anonymized people - share mutual interests.
Then, the participants will learn to read, interpret, briefly analyze, and visualize this graph in Gephi. This will include the selection of appropriate graph layout, learning about network filtering, computing basic statistics, and fine-tuning visual characteristics.
As a final outcome of the workshop, students will create and export their own image file visualizing the participant co-interest graph in their own unique style.
Finally, we are going to launch an online competition during the conference, where participants are encouraged to share their visuals online using the later-introduced hashtags. The creators of the most popular visuals will take home not only an awesome brand new book but the reputation of the entire community.
Key Topics
- Reviewing the fundamental components of a network visualization
- Parsing a graph object in Gephi
- Network layouts, filtering, and statistics
- Designing a network visualization
Target Audience
- Data Scientists and Analysts
- Data Engineering
- Machine Learning Engineers
- Data Analysts
- Managers of the above
Goals
Gain hands-on experience on how to visualize networks using Gephi.
Session outline:
- Brief introduction to the main components of a network visualization
- Parsing a prepared gexf file in Gephi
- Testing different network layouts
- Setting node size and color
- Computing network statistics
- Setting graph filters
- Fine-tuning and exporting the final visuals
Level
Beginner
Prerequisite Knowledge
Experience with any spreadsheet editor, like MS Excel