Repository with notebooks and complementary materials relate to the paper "LLMusic: Leveraging LLMs for Topic Modeling and Classification in Brazilian Funk Lyrics"
From the extend material, the dataset excerpt_analysis.xlsx was added. Each row in the dataset represents an excerpt from a song and its possible topic. Annotators (annotator_1, annotator_2, annotator_3) assign binary labels (1 for relevant, 0 for not relevant) to indicate whether the excerpt fits the given topic (topic_name). Multiple rows exist for each excerpt to evaluate different topics.