Use data to display the best "hot performances" in our mobile karaoke app
data analysis data structures data visualization algorithms data mining data entry
$390
top 6
20
submissions
20
solutions max
Small pic geandjeannie

The Smule community is most vibrant within our app Sing! Karaoke, which is essentially a social network for karaoke singers. The app allows people to sing their favorite songs and upload their performance to their profile on the Smule Network. All of the usual social mechanics are in place such as following other users, commenting, liking, etc and a social feed is generated for each user similar to other social network activity feeds.

The thing that makes our network unique is the collaborative element. Users can sing Solo songs, but can also create Duets or Group performances. This collaborative mechanic provides a brand new set of social interactions that creates interesting problems yet to be addressed by any other social network.

Within the app we have a section called "Hot Performances", which is devoted to surfacing interesting and unique performances. This is currently the only way users can discover content other than searching through the app, or sending each other links to performances. You can visualize this "Hot Performances" list on our homepage (www.smule.com) as well as in the app.

Deliverables

We currently surface content in "Hot Performances" using an algorithm that takes into account number of likes and listens over time, and we have two spots at the top of the list for Smule to feature Hand Picked performances. The hand picked performances stay, but the algorithmically surfaced performances rotate frequently. The problem is that some of our users have learned how to game our algorithm so it gets flooded with a ton of performances from a couple specific users who have no shame in gaming the system.

We are looking for new ways to surface content either algorithmically or by leveraging the community. Ideally we would be able to surface brand new talent, as well as allow our power users to get heard without having to resort to gaming the system.

1) Submit a description of new way(s) to surface content and how/where they could appear in the app, or on our website.

2) Describe how users could interact with this new module for content discovery.

    Submissions will be graded on the following criteria:
  • Meets Deliverables
  • Creativity
  • Clarity
Leaderboard
Top 3 share $300 Next 3 share $90
$100.00 Square pic 60 rofl New York University
$100.00 Square pic 60 face Clarkson University
$100.00 Square pic 60 me University of Texas at Austin
$30.00 Square pic 60 015365a   University of Maryland
$30.00 G Franklin W. Olin College of Engineering
$30.00 Square pic 60 323295 4562188926406 642670556 o University of Pennsylvania