Skip to content

Embedding Map Explorer

Run Fullscreen

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
<iframe src="https://dmccreary.github.io/search-microsims/sims/embedding-map-explorer/main.html"
        height="602px" width="100%" scrolling="no"></iframe>

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

  1. Browse the map - Explore subject clusters
  2. Click any point - See details and similar items
  3. Use search - Type to highlight matching items
  4. Filter by subject - Uncheck to hide categories
  5. Adjust point size - Use slider for dense areas
  6. 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:

  1. Assess the organization of MicroSim collections in embedding space
  2. Identify clusters, outliers, and semantic relationships
  3. 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)

  1. Start with all subjects visible
  2. Click a Physics MicroSim - what are its neighbors?
  3. Find "boundary" items between clusters
  4. Search for "wave" - where do matches appear?
  5. 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?

References