Modelling song popularity as a contagious process

Area of Science:

• Epidemiology
• Musicology
• Computational Social Science

Background:

• Popular songs are often described as 'viral' or 'contagious'.
• The spread of song preferences shares similarities with the transmission of infectious diseases.
• Understanding the mechanisms behind song popularity is crucial in the music industry.

Purpose of the Study:

• To investigate whether infectious disease transmission models can explain the spread of song preferences.
• To compare the effectiveness of the susceptible-infectious-recovered (SIR) epidemic model with a phenomenological spline model in fitting song download data.
• To explore genre-specific differences in song popularity dynamics.

Main Methods:

• Analysis of song download data from MixRadio (Great Britain, 2007-2014).
• Comparison of the standard susceptible-infectious-recovered (SIR) model against a spline model for fitting download time series.
• Fitting both models to simulated epidemic time series generated by the SIR model.

Main Results:

• The susceptible-infectious-recovered (SIR) model provided a better fit to song download time series compared to the spline model.
• The SIR model's superior performance in fitting song downloads mirrored its performance in fitting simulated epidemic data.
• Estimated SIR parameters suggested faster preference spread for Electronica music, potentially due to a more connected 'susceptible community'.

Conclusions:

• The social processes driving song popularity are analogous to infectious disease transmission.
• Epidemic models like SIR offer valuable insights into the dynamics of music consumption and viral trends.
• Genre-specific analysis using SIR parameters reveals distinct patterns in music preference spread.