EDUCATIONAL AI TRANSFORMATION
Intelligent Generation of Physical Education Teaching Video Assessment by Integrating Big Language Modeling and Stacking Algorithm
At present, the traditional physical education teaching mode can no longer meet the ever-changing teaching changes, and how to improve the quality of physical education teaching has become the focus of current research. To improve the quality of physical education teaching and enhance students' mastery of physical education knowledge points, a method for intelligent generation of physical education teaching video evaluation based on the fusion of big language model and stacking algorithm is proposed. The new method extracts semantics from physical education teaching videos through a large language model, and uses a stacking algorithm to evaluate and analyze the generated test questions. The research results indicate that the stacked algorithm model has significantly higher indicator values than traditional models in the evaluation of short answer and essay questions. After using the model in teaching validation analysis, the number of knowledge points mastered by students increased from 8 to 9, and the evaluation score increased from 4.2 points to 8.5 points. It can be seen that the new method can significantly improve the evaluation effect of sports teaching videos and increase students' mastery of teaching knowledge points. This has good guiding value for improving the quality of physical education teaching.
Authors: Yu Chen, Jiadong Zhu
Executive Impact: AI-Driven PE Teaching Enhancement
Our innovative approach significantly elevates the quality and effectiveness of physical education through intelligent video assessment, boosting student learning outcomes and streamlining evaluation processes.
0
SHORT ANSWER BLEU IMPROVEMENT
0
ESSAY QUESTION BLEU IMPROVEMENT
0
KNOWLEDGE POINTS MASTERED
0
AVERAGE EVALUATION SCORE INCREASE
Deep Analysis & Enterprise Applications
Select a topic to dive deeper, then explore the specific findings from the research, rebuilt as interactive, enterprise-focused modules.
Large Language Model for Semantic Extraction
The Big Language Model (LLM) acts as the core for semantic extraction and question generation. It transforms physical education video content into structured text using speech recognition and text parsing, creating a keyword system. Through prompt engineering, these keywords are combined into coherent teaching scenario descriptions and linked to a knowledge graph. This dynamic process, integrated with attribute configuration, automatically generates targeted assessment questions aligned with teaching progress and knowledge point weights. The system constructs a comprehensive teaching knowledge framework, enabling intelligent assessment generation based on video content.
Stacking Algorithm for Precise Assessment Scoring
To accurately score the generated assessment questions, a robust stacking algorithm is employed, utilizing the Bidirectional Encoder Representations from Transformers (BERT) model. This method retrieves new test question data, feeds it into BERT to extract entity information, and trains it to predict scores. Different feature-based test scores are then integrated to produce a final, comprehensive score. A critical evaluation feedback module within the stacking algorithm segments long texts, processes them through BERT for encoding, and feeds the predictive evaluations back to the LLM for iterative problem generation optimization, enhancing overall assessment accuracy and adaptability.
Validated Performance and Positive Educational Impact
Experimental results confirm the superior performance of the improved BERT model with stacking algorithm. In short-answer assessments, it achieved a maximum BLEU value of 203, significantly outperforming other models by 125 points. For explanatory questions, it reached a maximum BLEU value of 201, exceeding BiLSTM by 115 points. Furthermore, the model significantly enhanced student learning: knowledge points mastered increased from an average of 8 to 9, and average evaluation scores rose significantly from 4.2 to 8.5 points. This demonstrates the model's ability to improve teaching effectiveness, student understanding, and application of sports knowledge.
Intelligent Evaluation Generation Structure
Data Introduction
→
Model Introduction
→
Attribute Introduction
→
Introduction of Test Questions
| Metric | Traditional Models (e.g., BiLSTM) | Improved BERT Model |
|---|---|---|
| Short Answer BLEU (Max) | 78 | 203 |
| Essay Question BLEU (Max) | 86 | 201 |
0
Increased in Average Student Evaluation Scores
Enhanced Student Learning Validation
Challenge: Traditional methods led to lower knowledge mastery (avg. 8 points) and evaluation scores (avg. 4.2 points) in physical education videos.
Solution: Implemented the AI-driven assessment system, integrating LLM for semantic extraction and a stacking algorithm (BERT) for precise scoring and iterative problem generation.
Result: Post-implementation, students' knowledge point mastery increased from 8 to 9, and average evaluation scores rose significantly from 4.2 to 8.5 points, validating the model's effectiveness in improving teaching quality and learning outcomes.
Calculate Your Potential ROI
Estimate the impact AI-driven intelligent assessment could have on your organization.
Estimated Annual Savings
$0
Annual Hours Reclaimed
0
Your AI Implementation Roadmap
A structured approach to integrating AI-driven assessment into your physical education program for maximum impact.
Phase 1: Needs Analysis & Resource Preparation
Conduct a thorough needs assessment to identify specific challenges and objectives. Prepare video data, establish knowledge bases, and configure relevant teaching tasks and knowledge points.
Phase 2: Model Training & Customization
Train the large language model for semantic extraction and question generation, and fine-tune the stacking algorithm (BERT) for optimal assessment scoring. Customize attributes for dynamic question generation.
Phase 3: Assessment Generation & Application
Generate personalized assessment questions from physical education videos and apply them in teaching scenarios. Utilize the integrated system for real-time evaluation and feedback.
Phase 4: Effectiveness Evaluation & Iterative Optimization
Evaluate the impact on student learning outcomes and teaching effectiveness using key metrics. Continuously refine the model and question generation strategy based on feedback and performance data.
Ready to Transform PE Teaching?
Book a free consultation with our AI specialists to discuss how intelligent video assessment can elevate your educational outcomes.
Ready to Get Started?
Book Your Free Consultation.
Let's Discuss Your AI Strategy!
Lets Discuss Your Needs
Select a Date
Sun
Mon
Tue
Wed
Thu
Fri
Sat