This article addresses the challenges of emotion recognition in song lyrics, which often diverge from the overall song's sentiment, by proposing a hybrid annotation framework that optimizes alignment between humans and large language models (LLMs). The authors introduce a new sentence-level dataset to examine this alignment and highlight the inherent subjectivity of the task.
- A new sentence-level dataset of lyrics is created to study human-LLM alignment in annotation tasks.
- The study highlights the subjectivity of lyric annotation and the challenges arising from misalignment between human and LLM perspectives.
- A hybrid framework is presented that optimizes both human and LLM annotation by predicting potential misalignments.