System and method for finding mood-dependent top selling/rated lists
A system and method for determining a rank aggregation from a series of partial preferences is presented. A distribution is learned over preferences from partial preferences with sparse support. A computer receives a plurality of partial preferences selected from two or more preference lists. Weights are assigned to each of said plurality of partial preferences, resulting in multiple ranked lists.
1. A method for determining multiple rankings from a set of partial preferences comprising the steps of:
receiving, by a computer, a plurality of partial preferences reflecting the preferences of a population;
forming a plurality of preference lists, where each preference list is a collection of partial preferences collected from a sub population of said population;
learning sparse distribution over all possible ranked lists or preference lists that is consistent with the formed plurality of preference lists such that the resulting distribution assigns positive probability to the fewest possible ranked lists; and
processing said learned sparse distribution over ranked lists to produce multiple popular ranked lists by identifying the ranked list that have positive probability under the learned sparse distribution over ranked lists,
wherein each ranked list of said plurality of ranked lists is said to correspond to a mood or type of a sub population of said population.
2. The method of claim 1 , further comprising the step of selecting said plurality of partial preferences from said first preference list and said second preference list.
3. The method of claim 1 further comprising the step of assigning a score to each of said plurality of partial preferences.
4. The method of claim 1 further comprising the steps of:
comparing a single partial preference to said plurality of partial preferences; and
adjusting the weight of said first preference list and said second preference list.
5. The method of claim 4 , further comprising the step of selecting said single partial preference from said first preference list and said second preference list.
6. The method of claim 1 , further comprising the step of determining a partial preference from a preference ordered list of items.
7. The method of claim 1 , further comprising the step of determining a partial preference from a pair-wise comparison.
8. The method of claim 1 , further comprising the step of selecting said first preference list and said second preference list from a population of partial preferences.
9. The method of claim 8 , further comprising the step of identifying a sparsest distribution consistent with said population of partial preferences.
10. A system for determining multiple rankings from a set of partial preferences comprising a computer comprising a processor and a memory configured to perform the steps comprising:
receiving a plurality of partial preferences reflecting the preferences of a population;
forming a plurality of preference lists, where each preference list is a collection of partial preferences collected from a sub population of said population;
learning sparse distribution over all possible ranked lists or preference lists that is consistent with the formed plurality of preference lists such that the resulting distribution assigns positive probability to the fewest possible ranked lists; and
processing said learned sparse distribution over ranked lists to produce multiple popular ranked lists by identifying the ranked list that have positive probability under the learned sparse distribution over ranked lists,
wherein each ranked list of said plurality of ranked lists is said to correspond to a mood or type of a sub population of said population.