Advanced Data Analysis with KNIME
Discover how to extract insights from online news and social media data using advanced analytic techniques such as AI-based clustering and social network analysis. This hands-on tutorial guides you through real-world media analysis workflows in KNIME—from grouping similar articles to identifying key influencers on Twitter. Ideal for communication professionals, media analysts, and students looking to explore the power of data without writing code.
Description
In this course, you'll explore two powerful techniques to analyze media coverage on “AI and research”:
- Clustering techniques – to group similar news articles and uncover patterns in how topics are covered
- Social Network Analysis (SNA) – to extract interactions from Twitter and build visual graphs to identify key influencers and community structures
The entire process is built in KNIME, a free and open-source analytics platform that lets you work visually with data—no programming required.
What you’ll learn:
✅ The basic theory behind clustering and its role in unsupervised machine learning
✅ How to preprocess media datasets: normalize data, remove outliers, handle missing values
✅ How to configure and interpret clustering in KNIME (using the K-means algorithm)
✅ How to identify top-performing sources per cluster and assign meaningful labels (e.g., "mainstream," "institutional," "niche")
✅ The fundamentals of Social Network Analysis and how it applies to social media
✅ How to create a mention graph from Twitter data using KNIME's Network Mining tools
✅ How to identify influential actors, map communities, and measure engagement structures
Who it's for:
This tutorial is perfect for media analysts, communication researchers, students, and data science beginners interested in practical clustering applications using real media coverage data.
What's included:
- A step-by-step slide deck (with speaker notes)
- Visual KNIME workflow screenshots
- Real clustering results from a Meltwater export
- Interpretive examples and tips for real-world use
No programming skills required—just curiosity and a basic understanding of data analysis.