Embedding Map Explorer
About This MicroSim
This MicroSim provides an interactive exploration interface for a MicroSim collection organized by semantic similarity, allowing users to discover related content through visual navigation.
Iframe Embed Code
1 2 | |
Description
Map Features
| Feature | Purpose |
|---|---|
| Clusters | Subject-based groupings (Physics, Chemistry, etc.) |
| Point size | Quality score (larger = higher quality) |
| Search | Find specific MicroSims by title |
| Filters | Show/hide by subject |
| Similar list | Top 3 similar items for selected point |
How to Use
- Browse the map - Explore subject clusters
- Click any point - See details and similar items
- Use search - Type to highlight matching items
- Filter by subject - Uncheck to hide categories
- Adjust point size - Use slider for dense areas
- Toggle labels - Show all point names
Visual Elements
- Colored points: Subject-coded (Physics=blue, etc.)
- Point size: Quality score (80-95%)
- Search glow: Yellow highlight on matches
- Selection glow: Gold highlight on selected
- Details panel: Title, subject, area, quality, similar items
Learning Objectives
After using this MicroSim, students will be able to:
- Assess the organization of MicroSim collections in embedding space
- Identify clusters, outliers, and semantic relationships
- Navigate embedding visualizations for content discovery
Lesson Plan
Introduction (5 minutes)
- Explain embedding maps as "semantic landscapes"
- Each point is a MicroSim positioned by meaning
- Clusters form naturally from similar content
Guided Exploration (15 minutes)
- Start with all subjects visible
- Click a Physics MicroSim - what are its neighbors?
- Find "boundary" items between clusters
- Search for "wave" - where do matches appear?
- Filter to show only Math - observe cluster structure
Discovery Activity (10 minutes)
- Find an "outlier" - a point far from its cluster
- Find a "bridge" - a point connecting two subjects
- Identify subclusters within a main cluster
- Find the highest-quality item in each cluster
Discussion (5 minutes)
- How would this help a teacher find content?
- What does it mean when items cluster unexpectedly?
- How could this guide new content development?