Algorithms Discriminate Women?

Discrimination against women in the music industry is apparent. Misogynistic men have the power in the music industry, which leads to discrimination against women. Recent studies by Pompeu Fabra University (UPF) and the University of Utrecht (UU) found that music algorithms on Digital Streaming Platforms have a huge influence on the gender gap in the music industry.  The investigators tested a commonly used “music recommendation algorithm based on collaborative filtering”. They analysed 330,000 users listening behaviour over nine years, in which just 25% of musicians that were listened to were women.

 

Digital Streaming Platforms are mostly to blame as their exclusive music playlists consist of men. Research shows their algorithms favour men also subsequently advance gender disparities within the music industry. Male Artist continues to be vigorously elevated to music audiences, which lead to an increase in streams. As a result of this, they are more likely to appear at festivals. Especially with these Digital Streaming Platforms, sometimes sponsor festival line-ups. Algorithms won't assist in eradicating gender inequality in music until people in power choose to change it.

The algorithm produces a hierarchy of artists to recommend to the audience. The Investigators discovered the primary suggestion of female musicians comes in sixth or seventh position, compared to the men where they come first or second when it comes to music recommendations. According to the Investigators, "The bias in exposure comes from the way it generated recommendations”. This implies that female artist exposure is almost diminished based on the recommendations of the gender bias system in place. The Investigators said the algorithm learns as the audience listens to suggest music from the Digital Stream Platforms. This now creates a pattern of male artists being predominantly streamed; with the assistance of a reordered algorithm, listeners will change their conduct.

The Investigators put forward an alternative method that creates greater outreach for female artists and would comprise reordering the recommendation of songs. This would move a predefined number of female artists into high positions in playlists and recommendations to tackle the current gender bias. The outcome of this investigation showed the assistance of the reclassified algorithm Digital Streaming Platforms users would change their conduct and listen to a more female artist compared to the current state of the recommendation algorithm. The reclassified algorithm is based on machine learning and would urge a change in listening behaviour.