|
|
|
|
|
|
by physcab
5138 days ago
|
|
|
Congrats on the launch! I remember when we (I was a former groovesharker) started working on it back in November of 2010! Great to see that you stuck with the project through what I'm sure were many ups and downs :) I'm curious to know how you see the product benefiting people. What do you see people actually using it for? Right now it seems like a cute demonstration of a data mashup though not entirely factually correct. For example, if I search for "Coldplay" it says 85% affinity for Asian Indian and 82% Plastic Surgery and high Z-score for Hindi? How do you ensure the accuracy of this data? I think you need to display the results of this data differently to users. If its correct, people without statistics backgrounds will get confused because not everyone can interpret what a z-score means, nor will they understand how it is useful to them. |
|
|
We addressed some of your questions in the faq, especially different ways that artists can use Beluga: http://beluga.grooveshark.com/faq.html. Essentially we wanted to give artists a powerful - yet free - tool to let them understand their fan base.
As for data accuracy; although Beluga constantly surprises us with its insights, further investigation has always confirms those insights, at least as far as being solid statistics.
We toyed around with a lot of ways of displaying the data before coming to the final design. Oviously not everybody's going to know what a z-score is, but hopefully the interface is useful for those who don't and even more useful for those who do. We sprinkled helpful tooltips throughout the app to give some context to what's being displayed.
Thank you for your thoughtful feedback!