Algorithms that tailor music recommendations to listeners around the world are more likely to feature music created by male artists, preventing content from female artists from reaching large audiences, according to a new study.
Posted by the Association for Computer Machinery, according to the study, on average, male artists top the recommendations of streaming platforms, while women come in sixth or seventh position. In addition, the proportion of female songs on these lists is also much lower. than men’s music.
The music industry is already known for its gender disparity – musicians, producers and label managers continue to be under-represented; while the gender pay gap is widening. The researchers note that the algorithm reflects the bias existing in the study’s dataset, in which only 25% of artists were women, echoing the under-representation of female artists in the industry. As such, the algorithms’ sexist recommendations may further reinforce discrimination, especially with more and more users turning to music streaming platforms.
Algorithmic biases not only minimize the exposure female musicians receive in the short term, but have a far-reaching impact. Recommendation systems are scalable – algorithms learn from the choices people make and are designed to offer future suggestions based on what people choose. If users keep picking male artists at the top of these lists, the recommendation algorithm will show them more of that category. In the long run, algorithms can bury women’s content further down the lists, creating a vicious cycle of what some call a “feedback loop”.
Related to The Swaddle:
Men and Women See Different Jobs on Facebook Due to Algorithm Bias, Study Finds
Interestingly, since music listeners seem to rely heavily on recommendations, the same feedback loop can also help correct the bias – by simply recommending more female songs. In fact, the researchers tested whether reclassifying sexist recommendations can affect users’ listening behavior, and the results are positive. “The simulation of the feedback loop shows that the gender can be balanced in the long term by gradually increasing the exposure of women artists in the recommendations,” the study notes.
“… The world’s population is made up of 50% women. So it would be ridiculous if the system didn’t recommend them, ”said one of the many artists interviewed for the study, who acknowledged the gender bias prevalent in music recommendations.
An effort that ensures gender equity, however, may meet with strong opposition from listeners – they may resist such a shift that offers ‘positive disparate treatment’ or affirmative action that helps uplift minority sections. , in favor of women artists. To counter this, the study highlights the need to roll out positive anti-discrimination policies, but “only gradually until gender balance is achieved to avoid reactance”, as users are essential to achieve the goal of equitable gender representation.