University of Twente Student Theses
Creating and evaluating a lyrics generator specialized in rap lyrics with a high rhyme density
Wentink, M. (2023) Creating and evaluating a lyrics generator specialized in rap lyrics with a high rhyme density.
|
PDF
485kB |
| Abstract: | Current musical tools and education allow every day people to advance their musical skills, however, they do not offer creative stimulation. One aspect in which creative stimulation could be applied within music is lyrics. This research proposes an architecture for a rap lyrics generator that focuses on multisyllabic rhyme; rhyming on more than one syllable that can span multiple words. This is done by selecting rhyming words from an input sentence and filling the space between them with Bidirectional Encoder Representations from Transformers (BERT). A prototype for this architecture has been developed and evaluated which showed that generated lyrics score significantly lower than existing human-written lyrics on rhyme and coherence. However, the participants are, at times, also impressed by the output of the tool. This suggests that an improved implementation of the proposed program could be viable for generating rap lyrics that contain multisyllabic rhyme and coherence at a higher level. |
| Item Type: | Essay (Bachelor) |
| Faculty: | EEMCS: Electrical Engineering, Mathematics and Computer Science |
| Subject: | 54 computer science |
| Programme: | Computer Science BSc (56964) |
| Link to this item: | https://purl.utwente.nl/essays/94362 |
| Export this item as: | BibTeX EndNote HTML Citation Reference Manager |
Show download statistics for this publication
Show download statistics for this publicationRepository Staff Only: item control page