Topicmodel Examples

K-Means vs Leiden Comparison

Explore how two powerful clustering algorithms handle different types of data. K-Means clusters by geometric distance (position), while Leiden discovers groups based on network connectivity (links). See which one wins across 12 diverse datasets!

Discover Topics

Auto-discover Space vs Food from docs.txt

Use Existing Topics

Match docs to Astronomy vs Cooking

Tech Support 20

Hardware vs Software tickets

News Headlines 30

Sports vs Finance news

Customer Reviews 30

Positive vs Negative sentiment

Science Abstracts 45

Bio, Physics, Space

Smart Home 40

IoT Device Commands

Travel Modes 40

Air, Rail, Road, Sea

Global News Feed 150

Politics, Tech, Sports, Entertainment, Health

Regional Briefs 120

6 categories, medium density

World Wire 180

6 categories, heavy coverage

Mega Stream 240

6 categories, massive volume

Connection Builder

Raise the slider to create/add more links and see how clusters change.

Sparse Dense 0

K-Means Clustering

Leiden Algorithm

K-Means Modularity
0.000

Higher is better - measures cluster quality

Leiden Modularity
0.000

Higher is better - measures cluster quality

Interpretation

Select a dataset to see the comparison...