Embedded Files
Concept Map Generation
Concept Map Generation
ICALM uses natural language processing and large language models (LLMs) to turn flashcards into concept maps by extracting subject-relation-object triplets. These maps form the foundation of an evolving knowledge graph.
Chatbot + Reflection
Chatbot + Reflection
After each session, ICALM invites a short reflection and generates a concise AI summary. Comparing the two reveals blind spots and consolidates understanding—no long essays required, just targeted metacognition.
Recommender System
Recommender System
Based on what students have studied or shown interest in, ICALM suggests new flashcards by navigating the knowledge graph. It uses AI to find the next best topic to study, helping learners explore content more deeply and logically.
🧠 Random Selection (Active)
🧠 Random Selection (Active)
Presents random cards from the graph to reinforce memory and improve recall.
Students can ask for hints or explanations as they go.
🔗 MetaNode (Constructive)
🔗 MetaNode (Constructive)
Starts with highly connected (high entropy) concepts and expands outward.
Builds structured understanding from complex, interlinked ideas.
🎯 Interest-Based (Interactive)
🎯 Interest-Based (Interactive)
Students input a topic of interest.
ICALM matches them to related flashcards using AI similarity scoring.
What's new in the 2025 build?
Page updated
Report abuse