LGS-KT: Integrating logical and grammatical skills for effective programming knowledge tracing.
Xinjie Sun1, Qi Liu2, Kai Zhang3
1School of Computer Science, Liupanshui Normal University, Liupanshui, China; School of Computer Science and Technology, University of Science and Technology of China, Hefei, China; State Key Laboratory of Cognitive Intelligence, Hefei, China.
This study introduces a new model for assessing programming skills by analyzing code quality and student interactions. The Logical and Grammar Skills Knowledge Tracing (LGS-KT) model significantly improves the accuracy of predicting student performance in programming exercises.
Area of Science:
- Computer Science
- Educational Technology
- Artificial Intelligence
Background:
- Traditional knowledge tracing (KT) methods inadequately assess programming skills by often overlooking crucial behavioral data during the coding process.
- Existing approaches primarily rely on exercise outcomes, failing to capture the nuances of student programming development.
- There is a need for more comprehensive models that integrate diverse data sources to accurately measure programming proficiency.
Purpose of the Study:
- To develop and evaluate a novel Knowledge Tracing model, Logical and Grammar Skills Knowledge Tracing (LGS-KT), specifically designed for programming education.
- To enhance the assessment of students' programming skills by incorporating both logical and grammatical aspects.
- To improve the accuracy of predicting student outcomes in programming exercises.
Main Methods:
- Integrated static analysis and dynamic monitoring (CPU/memory usage) to evaluate code quality and student programming behavior.
- Developed a reweighted logical skill evolution graph by analyzing multiple student iterations on programming problems.
- Constructed a grammatical skills interaction graph based on knowledge concept similarity to model grammatical skill interactions.
Main Results:
- The LGS-KT model demonstrated superior performance in predicting student outcomes compared to existing methods.
- Significantly improved the accuracy of inferring students' programming grammatical skill states.
- The integrated approach provided a more thorough assessment of code quality and student development.
Conclusions:
- The LGS-KT model offers a more effective approach to assessing programming skills by integrating logical and grammatical dimensions.
- This research highlights the potential of advanced KT models in personalized programming education.
- The study provides open-source data and code to foster further research and innovation in educational technology for programming.
More Related Videos
Related Concept Videos
Trial and Error and Algorithm
Components of Language
Machines: Problem Solving I
The toggle clamp system is a machine structure consisting of movable, pin-connected multi-force members that form a stabilized system to transmit forces. The...
Machines: Problem Solving II
Piaget's Stage 3 of Cognitive Development
Conservation and Constancy of Quantity
A significant cognitive milestone in the...
Theorems of Pappus and Guldinus: Problem Solving


