An Expert System to Generate Domain-Specific Programs using Natural Language Processing

This record contains the updated supplementary and experimental materials associated with the study entitled “An Expert System to Generate Domain-Specific Programs using Natural Language Processing.” The archive includes experimental task descriptions, the results of 70 test cases from Experiment 8, statistical data and calculations comparing the proposed expert system with ChatGPT-5 and GitHub Copilot, and supplementary documentation covering the experimental evaluation, including the robustness assessment in Experiment 29. The materials are provided to support the transparency, verification, and reproducibility of the experimental procedures and results reported in the manuscript.