Saltar al contenido

What Music Should I Listen to? Let’s Ask the Weather-Based Music Selector!

Weather influences our emotions and we listen to music for its vibes…why not combine it all into a tailored playlist?


Photo by Victor Xok from Unsplash

I know I’m not the only one that has a hard time finding music to listen to.

It has to be the right one for the mood I’m in, otherwise it won’t feel right.

And that mood of ours is influenced by many things, one of which is the weather. You know how there’s a saying for it, “being under the weather”, but we don’t usually think about how much it can affect us.

Mood states do change with weather, so why not use this variable to find the right music for us?

How are we classifying music today?

We have multiple libraries of music on the web and it’s not always easy to navigate through them.

If you search for certain keywords like “chill music”, you’ll find a variety of playlists that don’t seem to have anything in common. You thought you were narrowing your search, but the results are only increasing your current confusion.

Today, personalized music recommendations use one of these two approaches (or both but shown separately):

  • Content-based filter: music suggestions based on the user’s previous selections and positive ratings from that list of content. For instance, when you see a tag under a song saying “84% match”, the number is based on your listening history on that platform.
  • Collaborative filter: peer-endorsed music that takes the user’s demographic group as a reference and tells the user their preferences to help him in his decision-making process. For example, when you see the tag “most listened songs in the US”, it’s telling you what others listen to more often.

Either of these filters is based on people’s listening behavior rather than the song’s features.

We need a better music classification tool that can help us find the right set of songs quickly. And on top of that, we should be constantly improving that classification system by feeding it with new inputs and helping it learn from previous selections so that it get to know our personal preferences.

Wouldn’t you feel nice to have someone (or something) know you so much that it always has the best option to suit your needs?

How are we influenced by weather?

You’ve certainly felt happier experiencing a sunny day after having several days of rain. It feels like a relief and everything around you seems joyful.

Even the bird’s song sounds like a praise to mother Earth!

But that’s just the tip of the iceberg.

A variety of weather variables are influencing our psyche, and we aren’t aware of them in the slightest.

Research in the 80s found that people tend to be less concentrated and more sleepy in a humid environment. Considering the diverse weather variables measured in this study (hours of sunshine, precipitation, temperature, wind direction, wind velocity, humidity, and barometric pressure), humidity was the most important one related to concentration.

So next time if you or your children are unable to focus, you know what to blame.

They also found that the hours of sunshine are related to people’s level of optimism and helping behavior. More hours of light bring more hope to this world.

Another study focused on the relationship between temperature and seasons on our psyche.

With higher temperatures, people show an increase in their mood, a better memory, and are more receptive to new information.

But that’s just during the spring. When temperatures increase too much, during the summer, people’s mood decreases significantly.

As Thanos would say: “Perfectly balanced, as all things should be”

Now to the weather-based music selector!


Photo by cottonbro from Pexels

These five software engineers from Bangladesh set an ambitious goal: create a mood-weather classification and recommendation system to enhance the user’s music listening experience.

Wouldn’t you want to remove the struggle of searching for the right music through endless lists of songs with a system like this?

First, they focused on 6 music features and their relationship with emotions. For instance, fast tempo (or fast-paced music) tends to evoke excitement, surprise, or fear. Another feature, pitch, is also related to specific emotions: high pitch to happiness or anger, and low pitch to sadness, solemnity, and sometimes pleasantness.

Then they turned their attention towards the weather. Based on the person’s location, they’ll get the weather conditions classified into 7 variables and add several emotional attributes to each of these features. For example, temperature includes emotions such as happy, angry, tense, distressed, bore, and tired, since previous research has shown that an increase in temperature can affect these mood states.

Once the combination of music features and weather conditions are associated with mood states, the playlist generator can create a specific playlist by relying solely on the current weather condition or have the user’s input regarding his current mood state.

If the user doesn’t interact with the system, then the playlist is entirely based on the current weather state. And if the user does state his mood, the software can combine both music features and weather variables related to that emotion and create a more tailored playlist.

The system also learns based on the user’s feedback and it updates the weight of these variables for that particular song so that next time its selection accuracy is much higher for that user.

And there you have it, the weather-based music selector!

The Takeaway

Emotions are the main connection between music and weather.

When we listen to music something inside of us is awakened. We like how music makes us feel.

We might not be aware of its toll on us, but the weather also influences our mood.

Once we try to combine those two, music features and weather conditions, we can create a set of songs suited for the mood we’re in, or the one we strive towards.

Plus, if we rate the way these two are mixed into our playlist, we can help improve its classification accuracy.

The result will be a better-suited playlist for our specific needs.

Finally, something can understand us, even if it’s just software. A program that can help us avoid the endless search for songs, and makes us feel the way we want to feel in that moment.

Music, emotions, and weather.

It’s just the beginning of a personalized music journey.

What’s the next thing we should add to this mix?


By Pavle Marinkovic on .

Are you curious about the world of sound and music? Learn how music can enhance a plant’s growth, the way sound changes our sense of taste, understand the music industry, and much more! Join my newsletter to embark on this journey of sound awareness.

If you were interested in this article, I recommend you follow this link: Audiobranding


Deja una respuesta

Tu dirección de correo electrónico no será publicada.

Translate »