Pandora May Have Opened The Box – But MusicGraph Has Built A Better One
Senzari’s MusicGraph API is a powerful toolset for music app developers that allows them to personalize their playlist and recommendation offerings beyond just taste – extending to a user’s activities and greater context in some very compelling ways.
MusicGraph combines a massively large music data set on one side – multi-billions of data points – with graph analytics and machine learning engines, social, location, and activity-based user data. The results are truly awesome in scope and they offer a real competitive advantage as developers vie to give consumers increasingly sophisticated and personalized music experiences.
Huge Music Data Set And Semantic Framework Combination
MusicGraph provides an impressive depth and breadth of music-related information. Its scalable semantic graph framework factors in genre and lyric data but goes further, including things like tempo,and influences. It’s based on machine listening algorithms that extract about 4,000 features and 200 megabytes of data per song. Now, multiply that by about 20 million songs – versus Pandora’s approximately 1 million songs – and you have a sense of the order of magnitude of information that MusicGraph provides.
The result is a stunning quantity of quality semantic knowledge about the songs that will eventually reach the user, so an app can make recommendations and playlists based on the real data instead of trial and error without limiting the variety of possible songs.
Real-time Personalization Data Adds Even More Value
MusicGraph also allows devs to tap into data on the other side of the equation in the form of user information including social metrics and trends, and even data from torrents. This provides greater contextual awareness of who the user is, where they are, and what they’re doing in real time – all factors that affect how they consume music and what works for them.
Awesome Use Case – Match Tempos To Real-time Activity Level
The possibilities of what this 3scale-powered API allows developers to achieve and users to experience can be pretty magical.
Here’s a use case: a developer wants their music app to be able to generate a highly specific “Running in Detroit” playlist. Easy enough – MusicGraph’s massive semantic graph structure produces an entire list of songs by artists from Detroit or that mention Detroit in the lyrics.
But the playlist can also respond to the user’s activity level based on information from the user’s phone, matching the songs’ rate of beats per minute to the speed of the runner. Warm up, peak workout and cool-down – all covered by the app.
There are so many possibilities. Start using MusicGraph, free at https://developer.musicgraph.com/.
Senzari’s products apply the latest machine learning and big data techniques to generate real-time semantic graphs, developing a broader understanding of the relationships between users and media. The MusicGraph platform, originally planned as a Pandora clone to deliver recommendations for Senzari’s music streaming service, Wahwah, has evolved significantly into a product of its own.