What question did this study set out to answer?

To develop and assess an automated grading and content-generation system for geotechnical engineering coursework using large language models and cosine similarity.

November 27, 2025Machine Learning and Data Science in Geotechnics

Large language models for automated grading in geotechnics

View Full Paper

Ask AI

Bookmark

Authors

ESEnrico Soranzo

Discussion

Member takes

Implication

Evaluation study demonstrates consistent automated grading and dynamic feedback in engineering students, indicating improved assessment efficiency and learning outcomes.

Key Points

To develop and assess an automated grading and content-generation system for geotechnical engineering coursework using large language models and cosine similarity.
Built a web-based evaluation platform using a generative pre-trained transformer model and cosine similarity to grade synthetic and real student answers against reference solutions in real time.
Fine-tuned large language models on geotechnical textbooks and applied retrieval-augmented generation along with data augmentation to grade anonymized student technical project reports.
The automated grading system delivered objective, consistent evaluation of technical reports and open-ended questions.
Immediate real-time feedback helped students effectively identify and close domain-specific knowledge gaps.
Domain-specific fine-tuning and retrieval-augmented generation produced accurate, relevant engineering questions and reference answers.

Cite This Study

Enrico Soranzo (2025) studied this question.

synapsesocial.com/papers/694031d32d562116f2906d3e https://doi.org/10.1108/mlag-01-2025-0001

View Full Paper

Ask AI

Bookmark

Implication

Evaluation study demonstrates consistent automated grading and dynamic feedback in engineering students, indicating improved assessment efficiency and learning outcomes.

Large language models for automated grading in geotechnics

Discussion

Member takes

Implication

Key Points

Cite This Study

Large language models for automated grading in geotechnics

Discussion

Member takes

Implication

Key Points

Cite This Study

Also Consider

Also Consider