Computational Analysis of Latin Music Songs Through Tokenization. Case of Female Artists and Reggaeton

The following research focuses on the analysis of the computational analysis of songs from female reggaeton artists. For this purpose, 641 songs from 12 singers were reviewed, the information obtained was extracted using a Python script with Selenium library, thus allowing access to the compilation of the lyrics, then to establish patterns and similarities, the R software was applied, complemented with the Voyant Tools tool. The results show that the songs interpreted by women contain an abundant lexicon, which in turn is augmented with anglicisms that make their legibility difficult. At the same time, the positioning of women in a dominant, decisive and outstanding role with respect to the actions involved in the typicality of urban-Caribbean music is demonstrated. In conclusion, women as singers in reggaeton have managed to take over a space homogenized by patriarchy, thus fostering a diversification of the musical entertainment industry with other profiles, identities and lyrics, thus building a diversification of this subculture.