Criteria for picking thumbed
I have a pandora like software system where the user can thumb up a song or thumb down a song.Application, recognized as chavah, is going to be silverlight + c, but this is the language and platform neutral question.
My software needs to choose a song good user's preferences.I need a good protocol for this.Then you decide a random element from the set of all scores, with its opportunity however weighed by its, um, burden.The fast and dirty approach i'd think of here rrs always to sum over all weights.Pick a random number small compared to that sum.Loop total songs until currentweight+songweight> randomnumber(Altogether different currentweight+=songweight)
Clearly one can make this arbitrarily more complex by http://www.milagrofilms.ca/bracelets-pandora.html introducing collaborative filtering: )
Let's say five songs.First two thumbed older, next Pandora Salehttp://www.milagrofilms.ca/ one thumbed over, two simple.
1:1.5, 2:1.5, 3:0.5, 4:1, 5:1
Cheap Pandora Bracelets The sum of the this is5.5.Now we pick Cheap Pandora Charms Canada a non-Selected number
Now let's find the song this random number is actually.
Beging with currentweight = 0.First song's dietary = 1.5.Currentweight+1.5 we carry on with, But develop CurrentWeight by this song's weight.
As a result currentweight = 1.5 at this point.Next song's extra fat:Anymore 1.5.Fortunately, currentweight+1.5 == 3 > 2.43789.Therefore we chose the second song!
What you do here is to practically pick a random spot on a line, but increase the"Terrain"On that line in which choose a song if the song is thumbed up.
Whether favored many repetitions or not basically depends on how strongly you increase/decrease the songs' weights.
Cooler: )Well i been feeling a bit, and something i don like about my approach is that it fairly slow, pushing you to loop over, in average sumofweights/2 songs each and every choice.If you can live with even numbered ratios from the weights, you could also just put the songs into a list and choose just just one just put the unthumbed ones twice(As an example), The upthumbed three times since the downthumbed once(This would make same distribution as in the example above, you just not quite as flexible if you like to change those weights)