Music recommendation systems – a shift from what you like to what you need?


To say that we listen to music because we enjoy it seems obviously true. From this comes the assumption that MIRs (Music Recommendation Systems) should be designed to recommend the users songs that they might like. But what if we were to ask a different question and start from a different formulation? What if we would consider music as an “aesthetic technology” (Krueger 2014) that listeners use as a tool to achieve certain mental states? What if the question would be how to give the user the best tools they need in order to best achieve some goal?

Recent studies have shown (Demetriou 2016), that while music is a common occurrence in daily life, it is rarely the sole focus of activity. Unlike in the not-so-distant past, technological advances have made music listening a private and ubiquitous phenomenon. Consider the run and the gym; the library and the study hall; the office and the commute. In none of the cases could we exhaust the purpose of music-listening by considering it as moments of “pure” aesthetic enjoyment. Rather, the headphones and the sounds (not always even musical) serve a purpose: for example to block out the surrounding noise and/or to facilitate a better focus on task at hand. Even in cases such as a student getting ready for Friday night at home, it would often be too simple to say that they are only listening to music because “they like it”. Obviously it is not necessarily false to say so, but they are perhaps more importantly trying to facilitate a certain mood (in order to feel upbeat and open, so that they would have a good time and so on).

This is not lost on music streaming services such as Spotify. There are playlists for “confidence boosting”, “the morning coffee”, “life sucks”, “feeling good”, and “coping with loss” to name but a few. These playlists seem to be (for the moment) hand-curated in terms of their universal (or at least country-wide) applicability. Going forward, it seems likely that unless insurmountable problems arise that the next step will be context- and individually sensitive playlists that use a dynamic understanding of the user to facilitate their goals. I predict that in the coming years we will see music services making much use of sensors and wearable technology to facilitate better sports performance or increase focus & productivity at work for example. Furthermore, if the services take note of findings in the psychology of music and music therapy, they may at least attempt to improve the psychological well-being of their users. For example, it has recently been reported (Carlson et al 2016) that the musical strategies certain types of people use to deal with negative emotional states may be maladaptive. If the service could identify such a person in such a situation, they could conceivably recommend alternative types of music. Whether or not such a recommendation would be welcomed by users is another question, not to mention the privacy issues regarding such profiling. Furthermore, it is unlikely we should recommend Spotify as surrogates for professional help. On the other hand, it is already surely the case that people do use music services for such purposes, and perhaps improving their capacity to do so would be in the end a positive development.

In conclusion, as assumptions about the role of music in everyday life shift, new markets and possibilities come into view. I would not be surprised to see in a few years’ time a service that explicitly markets itself as “aesthetic technology” specifically designed to modify and maintain certain psycho-physical states that are seen as beneficial given some goal.

Of course, that music could be used as a tool to improve human performance is nothing new. Try the recently unearthed secret training music for the East German olympic program next time you go out for a run. This particular piece is the second section of a 5k run at 156 BPM.



Krueger, J. (2014). Affordances and the musically extended mind. Frontiers in Psychology, 4. doi:10.3389/fpsyg.2013.01003

Demetriou A., Larson M., Liem C. (2016) Go with the Flow: When Listeners User Music As Technology., 17th International Society for Music Information Retrieval Conference.

Carlson, E., Saarikallio, S., Toiviainen, P., Bogert, B., Kliuchko, M., & Brattico, E. (2015). Maladaptive and adaptive emotion regulation through music: a behavioral and neuroimaging study of males and females. Frontiers in Human Neuroscience, 9. doi:10.3389/fnhum.2015.00466



1 Comment

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s