Personalized Learning Path Planning with Goal-Driven Learner State Modeling

2025-10-15

AI Summary

Personalized Learning Path Planning (PLPP) suffers from insufficient goal alignment. To address this, we propose Pxplore—a goal-driven framework that integrates reinforcement learning (RL) with large language models (LLMs) to construct a structured learner state representation and an automated reward function, thereby translating abstract learning objectives into computable signals for dynamic, goal-consistent path generation. Methodologically, we jointly train the LLM policy via supervised fine-tuning (SFT) and group-relative policy optimization (GRPO), underpinned by an education-specific decision architecture. Experiments demonstrate significant improvements in path coherence, personalization, and goal alignment; Pxplore has been deployed in a real-world learning platform, with code and datasets publicly released. Our core contribution is the first deep integration of LLM-driven RL into goal-aligned PLPP—enabling a paradigm shift from “experience-based recommendation” to “goal-verifiable planning.”

Technology Category

• Search and Optimization: Learning to Search
• Machine Learning: Large Multimodal Models (LMMs)
• Planning, Routing, and Scheduling: Planning with Language Models

Application Category

• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation
• Search and Retrieval-Augmented AI: Search Tool Learning with LLM: Teaching LLMs to invoke search and make use of retrieved information
• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems

Abstract

Personalized Learning Path Planning (PLPP) aims to design adaptive learning paths that align with individual goals. While large language models (LLMs) show potential in personalizing learning experiences, existing approaches often lack mechanisms for goal-aligned planning. We introduce Pxplore, a novel framework for PLPP that integrates a reinforcement-based training paradigm and an LLM-driven educational architecture. We design a structured learner state model and an automated reward function that transforms abstract objectives into computable signals. We train the policy combining supervised fine-tuning (SFT) and Group Relative Policy Optimization (GRPO), and deploy it within a real-world learning platform. Extensive experiments validate Pxplore's effectiveness in producing coherent, personalized, and goal-driven learning paths. We release our code and dataset to facilitate future research.

Problem

• Planning adaptive learning paths for individual goals
• Transforming abstract objectives into computable reward signals
• Developing goal-driven learner state modeling framework

Innovation

• Reinforcement-based training paradigm for personalized learning
• LLM-driven educational architecture with structured learner modeling
• Automated reward function transforming goals into computable signals