Person to person similarities based on media experiences

A computer-implemented method, comprising:

• under control of a computing device configured with executable instructions, performing the following:
• selecting, from a population of users, a first sample user who experiences a first set of media items;
• defining a first individual user-based cluster according to the first set of media items experienced by the first sample user;
• selecting, from the population of users, a second sample user who experiences a second set of media items;
• defining a second individual user-based cluster according to the second set of media items experienced by the second sample user;
• and
• repeating the selecting and defining to form C individual user-based clusters, where C is a positive integer greater than two;
• and
• under control of a computing device configured with executable instructions, performing the following in substantially real time:
• receiving a query from a non-sample user;
• and
• determining a media experience similarity between the non-sample user and one or more other non-sample users with respect to the media items the non-sample users have experienced, the media experience similarity between the non-sample user and each of the other non-sample users being derived in part based on a comparison of a first indicator and a second indicator, the first indicator indicative of how similar a media experience of the non-sample user is to a media experience associated with each of the individual user-based clusters and the second indicator indicative of how similar a media experience of the other non-sample user is to a media experience associated with each of individual user-based cluster.

A non-transitory computer-readable media storing computer-readable instructions that, when executed, instruct a processor to perform acts comprising:

• forming multiple individual user-based clusters according to behaviors exhibited by individual sample users on whom each of the clusters are based;
• mapping the user-based clusters as orthogonal vectors in a multi\-dimensional space;
• placing a first non-sample user into the multi-dimensional space according to how the media items experienced by the first non-sample user compares with the media items in each of the individual user-based clusters;
• placing a second non-sample user into the multi-dimensional space according to how the media items experienced by the second non-sample user compares with the media items in each of the individual user-based clusters;
• and
• computing a similarity metric between the first and second non-sample users as a function of a distance from the first non-sample user to the second non-sample user within the multi-dimensional space.