CIB Conferences
Abstract
This study investigates the potential of large language models (LLMs) to enhance feedback generation for written assignments, with a focus on Leadership in Energy and Environmental Design (LEED) projects. Through a systematized review of 70 studies, we identify key trends, including a predominance of structured (52.63%) and argumentative inputs (21.05%) and iterative feedback mechanisms (64.91%), which align with LEED’s criteria-driven workflows. However, gaps remain in addressing technical inputs (3.51%) and scalable automated feedback systems tailored to LEED-specific needs. To address these challenges, a conceptual framework is proposed, emphasizing advanced natural language understanding, real-time iterative feedback, and automated validation systems to support the structured and technical demands of LEED assignments. While this framework offers potential for improving feedback systems in high-stakes educational contexts, further experimental validation is necessary to assess its effectiveness in practice.
The paper will be presented:
Online
Primary U.N. Sustainable Development Goals (SDG)
Sustainable Cities and Communities - - Make cities and human settlements inclusive, safe, resilient and sustainable
Secondary U.N. Sustainable Development Goals (SDG)
Quality Education - - Ensure inclusive and equitable quality education and promote lifelong learning opportunities for all
Primary CIB Task Group OR Working commission
W116 – Smart and Sustainable Built Environments
Secondary CIB Task Group OR Working commission
W089 – Education in the Built Environment
Recommended Citation
Yang, Zhenlin; Castelblanco, Gabriel; Cruz-Castro, Laura; and Multani, Maryam K.
(2025)
"Proposed Framework for LLM-Based Feedback in LEED Assignments,"
CIB Conferences: Vol. 1
Article 103.
DOI: https://doi.org/10.7771/3067-4883.2036