ECNUClaw is an open-source framework by Zhou, Li & Zhang (2026) for building learner-profiled intelligent study companions in K-12 education. The system maintains a five-dimension learner profile — cognitive, behavioral, emotional, metacognitive, and contextual — by extracting signals from student-companion dialogues at each conversational turn.

ECNUClaw: K-12 Personalized Study Companion

Core Contribution

ECNUClaw is an open-source framework by Zhou, Li & Zhang (2026) for building learner-profiled intelligent study companions in K-12 education. The system maintains a five-dimension learner profile — cognitive, behavioral, emotional, metacognitive, and contextual — by extracting signals from student-companion dialogues at each conversational turn.

How It Works

The system draws on three theoretical strands from Chinese educational technology literature:

1. Digital Portrait Three-Layer Framework (Zhang) — for learner assessment

2. Education Brain model — for educational system architecture

3. Human-AI Collaborative IQ — for companion design philosophy

Profile updates feed into an adaptive strategy engine that adjusts guidance intensity, encouragement frequency, and Bloom's taxonomy scaffolding in real time. The framework supports seven Chinese LLM providers through a unified OpenAI-compatible adapter layer.

Significance for AIED

ECNUClaw advances the field of Personalized Learning by operationalizing real-time learner profiling within dialogue-based tutoring. Unlike static learner models in Adaptive Learning, ECNUClaw updates profiles turn-by-turn, enabling genuinely responsive Intelligent Tutoring at scale. The five-dimensional profile connects to work on Metacognition (metacognitive dimension), Self Regulated Learning (contextual dimension), and Affective Tutoring (emotional dimension). The system's grounding in Chinese educational frameworks also extends the geographic scope beyond Western-centric Stanford Evidence Base AI K12 2026 research.

Open Questions

How does turn-by-turn profiling compare to LLM Student Modeling Memory approaches using persistent memory architectures?

Can the framework generalize beyond Chinese LLM providers and K-12 contexts?

What are the privacy implications of five-dimensional profiling, especially for minors in Regulation contexts?

Connected Concepts

Personalized Learning

Adaptive Learning

Intelligent Tutoring

Metacognition

Self Regulated Learning

Affective Tutoring

Regulation

Connected Articles

Stanford Evidence Base AI K12 2026

LLM Student Modeling Memory

Citation

Zhang, A.Y.Z.J.L.Z., TUDY, E.L.A.L.E.R.I.N.S., OMPANION, C., DUCATION, F.R.F.K.P.E.E., REPRINT, A.P., 1,2, Y.Z.J.L.A.Z.Z., & Normal, A.E.L.E.C. (2026). ECNUClaw: A Learner-Profiled Intelligent Study Companion Framework for K-12 Personalized Education