Recommendations
Not for Geeks Only
The instant recommendations given by MyStrands is a popular feature. Popular and, for many, somewhat mysterious as well. We are often asked how recommendations work and as you, the community of music lovers, are providing us with the information to make them, we are only too happy to answer!
It Takes a Village
There are various recommendation technologies in use today. The fancy name for the one most similar to how we do it at MyStrands is collaborative filtering, defined by Wikipedia to be a "method of making automatic predictions (filtering) about the interests of a user by collecting taste information from many users (collaborating)". In other words, our software gets to know a lot and (we mean a lot) of people in terms of the music they like. Are you playing "Drops of Jupiter" by Train (again!)? MyStrands happen to know that many fans of that track also like to play "You're Beautiful" by James Blunt. So if you haven't yet checked out the U.K. pop-rock phenomenon, then perhaps you should.
Getting to Know You
The question remains, how do we learn about people's musical taste? As you know, or soon will as you explore MyStrands.com, what you play counts®. Go to a member's page and you'll see his or her favorite artists, top tracks, the playlists they've made and so forth. In other words, their musical taste as tallied by MyStrands. But we don't stop there. The zillions of public pages on the World Wide Web are nothing less than a repository for the musical tastes of millions of rockers, long hairs, teeny boppers, punks, goths, opera fiends, flaca fans, country connoisseurs, psychedelic rangers, funk lovers, ..., well, you get the idea. MyStrands has analyzed over one million playlists alone. We surf, we learn. It's really that simple, although the braniacs around here call it data mining.
Big Brother is Not Watching
While all that poking around the Web might seem a bit intrusive, be assured that our recommender doesn't keep any personal information. In looking at a playlist we might see that a Beatles fan also likes The Rolling Stones. We give that information to the recommender, but we neither look for nor include data about the person behind it. We don't have to; taste is taste.
Filters: Just the Way You Like It
Our recommender is not a one-size-fits-all operation. We also have filtering technology that allows you to customize your recommendations based on qualities such as genre and popularity. Maybe you're listening to "Flake" by Jack Johnson, and our recommendations are so spot on that you already know them all. By opening the filter drawer on the MyStrands recommendations window (just click the gear), you can request recommendations of less well-known songs, giving you "Pink Noise" by Beck, instead.
So, don't wait! Download MyStrands, keep listening, and start discovering!
