📄 Full text: arXiv:2604.26962 · local · arXiv:2409.03512 · local · arXiv:2604.13074 · local
Tailoring educational experiences to individual learner profiles, including prior knowledge, learning pace, preferences, and affective states. AI enables personalization at scale, though the gap between system personalization and learner-perceived personalization remains an open measurement challenge.
Architectures for AI-Driven Personalization
Longitudinal Memory (PersonaVLM → Education)
Nie et al. (2026) developed a multimodal long-term memory architecture (PersonaVLM) that maintains persona consistency across interactions. Mapped to education, this enables tutoring systems that remember a learner's misconceptions, preferred explanations, and progress history across sessions—addressing a critical deficit in stateless chatbot tutors.
Agent-Native Personalization Substrate (DeepTutor)
Ma et al. (2026) design every DeepTutor feature to share a common personalization substrate, rather than bolting personalization onto reactive tools. This architecture ensures cross-modality coherence: the same learner profile drives problem solving, question generation, and collaborative writing.
Multi-Agent Social Personalization (MAIC)
Yu et al. (2024) personalize not only content but social context. Classmate archetypes (Class Clown, Deep Thinker, Note Taker, Inquisitive Mind) create varied peer-learning dynamics matched to individual learner needs.
Measurement Challenges
- System vs. perceived personalization — A system can adapt without the learner feeling recognized
- Longitudinal validity — Personalization benefits may decay if profiles become stale or overfit
- Equity risks — Over-personalization can strand learners in low-expectation tracks
Related Pages
- kt4eqg-personalized-question-generation — 7 of 8 papers in May 28 scan
- learnmate2-llm-adaptive-learning — LLM-powered personalized study plans and adaptive activities
- taklif-ai-interest-based-personalized-assignments — Interest-based personalization using LLMs
- computational-thinking-ai-agent-creation — Optimal Development Zone effect for personalized instruction
- temporal-smoothness-debiased-kt — Unbiased mastery estimates for reliable adaptive recommendations
- conversational-ai-tutors-framework — GenAI enables dynamic content generation for personalized tutoring
- self-referential-l2-writing-llm-assessment — Profile-based assessment aligns with personalized learning goals
- ecnuclaw-k12-personalized-companion — Open-source K-12 framework with five-dimension turn-by-turn learner profiling
- cognitive-agent-compilation — Explicit learner models in CAC enable inspectable, editable personalization
- ai-adult-learning-design — 19 design guidelines for AI technologies supporting adult learners
- ai-learning-companions-framework — three-foundation framework for AI learning companions prioritizing durable learning over performance
- adaptive-learning-systems — The technical systems that deliver personalized experiences
- llm-student-modeling-memory — Longitudinal memory for maintaining learner profiles
- knowledge-tracing-irt — Models that infer what a learner knows in real time
- tutoring-specific-vs-general-ai — Design trade-offs when personalization is the goal
- agentic-workflows-education — Agentic paradigms for orchestrating personalized tutoring
- a4l-analytics-pipeline — A4L modular analytics pipeline for cross-domain educational AI data
- history-aware-student-simulation — History-conditioned student simulation with RL-optimized profiles for tutoring
- instructional-guidance-genai-learning — Five-step GLT-based prompting improves personalized AI learning- dynamic-skill-matching-capstone-teams -- LLM-driven team formation for capstone projects outperforms CATME Team-Maker in skill coverage and preference alignment
- learnopt-exam-cognitive-structure -- Standardized exams have stable latent cognitive structures recoverable via LLM-tagged question analysis and knapsack optimization
- edtech-design-time-generative-ui -- Design-time GenUI with card-based semantic units embeds UDL and eliminates per-learner inference costs
- self-efficacy-tutoring-learning — Self-Efficacy and Favorability Shape Learning from Tutoring
- lecturaagents-multi-agent-teaching — LecturaAgents
- code-anchor-multi-view-visualization -- Code as Anchor, Memory and Metaphor as Support: Learner Experiences with Multi-View Visualizations
- learning-to-prompt-adaptive-tutoring -- Learning to Prompt: Improving Student Engagement with Adaptive LLM-based High-School Tutoring
- adaptive-pretesting-retention -- Adaptive AI pretesting boosts initial learning, but long-term retention over 7 weeks depends on subsequent retrieval practice — learner-directed AI study underperforms structured retrieval.
- bridging-instructional-design-framework-math -- Proposes operationalizing learning theories as metadata dimensions for teacher-support systems in mathematics education, focusing on conceptual structure of content.
- skill-acquisition-without-temporal-info -- A neural framework predicts next-skill acquisition from snapshot CDM data using pseudo-temporal ordering via inclusion relations among skill sets.
- courseblueprint-adaptive-video-generation -- Explicit typed instructional contracts (scaffolding, adaptive controller, engagement contract, slide override) drastically improve pedagogical video quality over fluent but PCK-absent text-to-video.
- ai-interlocutor-l2-spoken-dialogue -- AI dialogue in L2 practice resembles 'supported monologue' with fewer turns but greater short-term linguistic uptake; human dialogue is faster and more balanced.
- automated-presentation-coaching — Personalized feedback in presentation coaching
- empathy-coaching-chatbot — empathy effects in coaching chatbots
- elevate-genai-virtual-tutors — personalized learning approaches
- pattern-kc-programming-recommendation — Pattern-based KC programming recommendation
- ai-learning-assistants-higher-ed-large-scale — Presents a large-scale descriptive analysis of an AI learning assistant (Syntea) using obj…
- learning-behavior-background-advantage-ai-ed — Personalization design (2026-07-14)
- llm-tts-dialogue-lesson-generation - adaptive lesson generation
- kutti-ai-voice-first-learning-companion — offline-first ASR extends personalization to low-resource settings
- student-math-competence-clustering — 119K students across 13 UK exams: overall ability dominates skill clusters; simp...
- ai-learning-tools-engineering-education-needs — 206 engineering students: AI chatbots give greatest benefit as relief from compe...
- hybrid-human-ai-tutoring-differentiated — System-level differentiation as personalization
- llm-children-reading-story-generation — Compact LLMs for difficulty-controlled reading stories
📎 59 other pages tagged personalized-learning
- A Semi-Automated System for Generating Dialogue-Based TTS Lessons Using Large Language Models: An Exploratory Study of Educational Potential
- A Survey of Automated Presentation Coaching: Systems, Methods, and Open Challenges
- AdaPT: Adaptive Lesson Plan Transformer for Cross-Regional and Differentiated Instruction
- Adaptive Learning
- Adaptive Learning Systems
- AI as a Partner in Learning about, Doing, and Engaging with Science: Vigilance as the Key to Productive Augmentation
- AI Coaching for Accelerating Human Skill Development with Reinforcement Learning
- AI-Generated Traces for Novice Programmers: Learning Effects and Learner Differences in a Multi-Institutional Study
- AI-Integrated Learning Management System for Middle School: A Longitudinal Study of Learning Outcomes
- Archetypes or ability? Clustering for modelling student mathematical competence
- Automated Recommendation of Programming Learning Content Using Pattern-based Knowledge Components
- Awareness of Technological Isomorphism: AI in Elementary Math
- Beyond Rephrasing: Book-Level Organization Improves Synthetic Textbook Data for Mid-Training
- Beyond Tool Adoption: A Practical Five-Stage Developmental Continuum for AI Literacy in Higher Education
- Building AI Companions that Prioritise Learning over Performance
- Cognitive Agent Compilation for Explicit Problem Solver Modeling
- Computational Thinking Development in AI Agent Creation: A Mixed-Methods Study
- CourseBlueprint: A Structured Pipeline for Adaptive Pedagogical Video Generation Grounded in Course Corpora
- Designing Needs- and Attention-Aware AI Learning Tools for Engineering Education: Insights from Psychological Outcomes
- ECNUClaw: A Learner-Profiled Intelligent Study Companion Framework for K-12 Personalized Education
- ELEVATE: Designing Human-Centered GenAI Virtual Tutors for Scalable and Inclusive Education
- Engagement Intensity as a Learner-Modeling Signal for Adaptive AI Ethics Instruction
- Fair and explainable educational recommendations with a hybrid Graph-GRU framework
- Feedback Loop
- Gen-AI-tecture: using generative AI to support architectural students in design tasks
- Generalizing a Highly Configurable Analytics Pipeline to Replicate and Support Educational Research Across Multiple Domains
- Generativism: Toward a Learning Theory for the Age of Generative Artificial Intelligence
- Guidelines for Designing AI Technologies to Support Adult Learning
- IKS-Instruct: A 24,000-Example Multilingual Dataset for Teaching Language Models Indian Knowledge Systems
- Improving Capstone Team Outcomes through Dynamic Skill Matching and Preference Alignment
- Improving Hybrid Human-AI Tutoring by Differentiating Human Tutor Roles Based on Student Needs
- Interpretable Difficulty-Aware Knowledge Tracing in Tutor-Student Dialogues
- Interpretable Knowledge Tracing via IRT
- Invisible Impact of Empathy on Behavioral Change: Isolating the Effect of Empathy in Long-term Physical Activity Coaching Chatbot Interactions
- KT4EQG: Personalized Exercise Question Generation via Knowledge Tracing
- Learning behavior accounts for background-related advantage in AI-assisted education
- Learning to Prompt: Improving Student Engagement with Adaptive LLM-based High-School Tutoring
- LearnMate^2: Design and Evaluation of an LLM-powered Personalized and Adaptive Support System for Online Learning
- LearnOpt: Recovering the Latent Cognitive Structure of Standardized Examinations via Knowledge Graphs and Constrained Optimization
- LecturaAgents: A Multi-Agent Framework for Adaptive Personalized AI-Assisted Learning and Embodied Teaching
- Lifelong Learning and AI
- LLM Student Modeling and Long-Term Memory Architecture
- ParaTutor: LLM Mediated Parent Child Tutoring through Role Separated Scaffolding Interface in Real Time
- PersonaVLM: Long-Term Personalization for AI Tutors
- Role of Instructional Guidance in Generative AI-Assisted Learning
- Self-Efficacy and Favorability Shape Learning from Tutoring Systems and Paper Practice
- Self-Regulated Learning
- Special-R1: Reinforcement Learning for Special Education — Aligning LLM Tutors to Diverse Learners through Disability-Adaptive Training
- Taklif.AI: LLM-Powered Platform for Interest-Based Personalized College Assignments
- TeachBench - Evaluating LLM Teaching Ability
- Temporal Smoothness Doubly Robust Learning for Debiased Knowledge Tracing
- The Missing Layer: Why EdTech Needs Design-Time Generative UI, Not Just Runtime Personalization
- The Path to Conversational AI Tutors: Integrating Tutoring Best Practices and Targeted Technologies to Produce Scalable AI Agents
- Tutoring-Specific vs. General-Purpose AI in Education
- Using AI-based Learning Assistants in Higher Education: A Large-Scale Descriptive Analysis
- Visualizing Engineering Fundamentals: Design of Mixed Reality and Physical Toolkits for Effective Learning
- What Changes When the Interlocutor Is an AI? Interactional Fluency and Linguistic Uptake in L2 Spoken Dialogue
- What Don't You Understand? Using Large Language Models to Identify and Characterize Student Misconceptions About Challenging Topics
- Who Am I? History-Aware Profiles for Student Simulation in Tutoring Dialogues