Spotify is continuing to tweak its algorithm, this time with the addition of patented "nostalgia metrics." According to a new US Patent & Trademark Office application, the Stockholm-based streaming giant intends to measure nostalgia through a combination of user and behavioral data.
The patent describes a system whereby listeners will be served tracks they find nostalgic based on their assignment into backend demographic groups. Spotify determines the demographic of a user based on a variety of factors, including age, geographic location, and gender, among other criteria.
With that data in mind, Spotify simultaneously begins identifying songs the user may find nostalgic based upon the year the song was released and the user's listening history. The algorithm also recognizes that people may engage with the same piece of music at different points and contexts in their lives, and not exclusively to invoke a sense of nostalgia. However, as Spotify continues to scale, listeners can expect the company's already complex predictive algorithms to become even more refined.
Many listeners have already discovered Spotify's ability to create a personalized feeling of nostalgia from the platform's Time Capsule playlist, which tailors uniquely tailors content to each individual user.
Spotify's algorithm has been successful in keeping its 300 million monthly active listeners engaged by making determinations about what the user prefers based on past behavior. As the platform continues to mature along with its audience, the company clearly is not shying away from finding new ways to apply its signature formula.
Check out the patent filing here.
Source: Digital Music News