Data and Code for «AI-Powered Modeling of Thematic Plasticity and Symbolic Recurrence in Lorca's Poetry»

This repository contains the computational tools, datasets, and supplementary materials used in the paper "AI-Powered Modeling of Thematic Plasticity and Symbolic Recurrence in Lorca's Poetry" (Digital Scholarship in the Humanities, 2026). The analysis covers the complete poetic works of Federico García Lorca (19 collections, 390 poems, 8,565 verses) across four creative periods. It combines lexical metrics, LDA topic modeling (Gensim), and LLM-assisted semantic interpretation (GPT-4o) to identify characteristic isotopies and thematic plasticity in Lorca's corpus. The repository includes the full analysis pipeline (Python/Colab), the input dataset, interactive LDA visualizations (pyLDAvis), and the LLM prompts and raw outputs used for topic interpretation. For replication instructions, refer to README.md.