Comment 1
Comment 2
Reference Information:
Title: Rush: Repeated Recommendations on Mobile Devices
Authors: Dominikus Baur, Sebastian Boring, Andreas Butz
Venue: IUI
Summary:
This paper described Rush, a recommendation based system for repeated item selection from large sets of data. The Rush system uses a touchscreen mobile device to find recommendations for similar items to the selected one based on either the "top five most similar items, five random selections from the list of similar items [or] a hybrid version of the two." What the system does in layman's terms is find similar songs to the one you are currently thinking of adding to a playlist. If you select a song from this recommendation, it will provide additional recommendations based on that song. This continues until the user is satisfied and adds the currently selected song to the playlist. Songs were considered similar based on Last.FM data.
Discussion:
This is a technology that I really wish I had. I often feel in the mood for a certain genre of song and want to create a playlist of these songs, but am frustrated with the time it takes to select them all. Even if the recommendations are simply ballpark estimates I would still find this to be an incredibly useful technology to have.
I see where the app would appeal to you for looking at songs but if I want to listen to a song I either want random or I want something specific I don't want to have to guess at similar songs until I find the correct one. None the less the app does seem useful.
ReplyDeleteI want to know what other applications the program could be applied to. I mean, a lot of research regarding intelligent suggestions use music as an initial application. What else is this good for? Besides other entertainment applications, like movies.
ReplyDeleteI know what you mean. When I'm bangin' seven gram rocks and finishin' 'em because that's how I roll, I do not have time to sift out all the Tay Zonday from my playlist.
ReplyDelete