Universal system and method for representing and predicting human behavior
First Claim
1. A computer-based service system comprising:
- a database that contains affinity data of a plurality of users;
a computer server configured to generate a subject affinity vector representation of a subject and an object affinity vector representation of an object, and to produce an object recommendation for delivery to an external application, the produced object recommendation based on the subject affinity vector, the object affinity vector, and the affinity data;
wherein the subject affinity vector and the object affinity vector each have a respective number of dimensions; and
wherein the predicted affinity of the subject to the object is calculated by matching the subject affinity vector to the object affinity vector; and
wherein the computer server is further configured to generate the subject affinity vector and the object affinity vector by producing initial subject affinity vectors and initial object affinity vectors having respective initial dimensions, to determine predicted affinity values based on the initial subject affinity vectors, initial object affinity vectors, and a subset of the affinity data, and to calculate a cost function; and
wherein the computer server iteratively increases the dimensions of the generated subject affinity and object affinity vectors, and recalculates the cost function based on the differences between the predicted affinity values and actual affinity values, until the cost function reaches a predetermined value, and wherein the actual affinity values are based on user input.
8 Assignments
0 Petitions
Accused Products
Abstract
A system and method is disclosed for profiling subjects and objects based on subjects'"'"' responses to various objects for purposes of determining and presenting the objects most likely to generate the most positive response from each visitor. Object ratings, such as aesthetic response, preference, interest, or relevancy, are explicitly submitted by subjects or derived implicitly from visitor interactions with the objects. Objects include movies, books, songs, commercial products, news articles, advertisements or any other type of content or physical item. A profiling engine processes the ratings information and generates compact profiles of each subject and object based on the similarities and differences in affinities between the group of subjects and the group of objects. A recommendation engine then generates recommendations to a subject based on similarity between the subject and object profiles. The recommendation engine can also match subjects to other subjects and objects to other objects. The recommendation engine can also predict affinity across object catalogs and across time. Additionally, the object profiles can be clustered to create behavioral object categories. The system has application in personalization, behavioral targeting, Internet retailing and interactive radio, to name but a few applications.
-
Citations
29 Claims
-
1. A computer-based service system comprising:
-
a database that contains affinity data of a plurality of users; a computer server configured to generate a subject affinity vector representation of a subject and an object affinity vector representation of an object, and to produce an object recommendation for delivery to an external application, the produced object recommendation based on the subject affinity vector, the object affinity vector, and the affinity data; wherein the subject affinity vector and the object affinity vector each have a respective number of dimensions; and wherein the predicted affinity of the subject to the object is calculated by matching the subject affinity vector to the object affinity vector; and wherein the computer server is further configured to generate the subject affinity vector and the object affinity vector by producing initial subject affinity vectors and initial object affinity vectors having respective initial dimensions, to determine predicted affinity values based on the initial subject affinity vectors, initial object affinity vectors, and a subset of the affinity data, and to calculate a cost function; and wherein the computer server iteratively increases the dimensions of the generated subject affinity and object affinity vectors, and recalculates the cost function based on the differences between the predicted affinity values and actual affinity values, until the cost function reaches a predetermined value, and wherein the actual affinity values are based on user input. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15)
-
-
16. A computer-implemented method of producing an object recommendation to an external application, the method comprising;
-
generating, at a server computer, a subject affinity vector representation of a subject and an object affinity vector representation of an object; producing an object recommendation for delivery from the server computer to an external application, the object recommendation based on the subject affinity vector, the object affinity vector, and affinity data of a database of a plurality of users; wherein the subject affinity vector and the object affinity vector each have a respective number of dimensions, the predicted affinity of the subject to the object is calculated by matching the subject affinity vector to the object affinity vector, and the computer server is further configured to generate the subject affinity vector and the object affinity vector by producing initial subject affinity vectors and initial object affinity vectors having respective initial dimensions, determining predicted affinity values based on the initial subject affinity vectors, initial object affinity vectors, and a subset of the affinity data, and calculating a cost function, and wherein the computer server iteratively increases the dimensions of the generated subject affinity and object affinity vectors, and calculates the cost function based on the differences between predicted affinity values and actual affinity values, until the cost function reaches a predetermined value, wherein the actual affinity values are based on user input. - View Dependent Claims (17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28)
-
-
29. A computer-based service system comprising:
-
a database that contains affinity data of a plurality of users; a computer server configured to generate a subject affinity vector representation of a subject and an object affinity vector representation of an object, and to produce an object recommendation for delivery to an external application, the object recommendation based on the subject affinity vector, the object affinity vector, and the affinity data; wherein the subject affinity vector and the object affinity vector each have a respective number of dimensions, the predicted affinity of the subject to the object is calculated by matching the subject affinity vector to the object affinity vector, and the computer server is further configured to generate an initial subject affinity vector having an initial number of dimensions and to generate an initial object affinity vector having the initial number of dimensions, based on a subset of the affinity data, and is further configured to modify the initial subject affinity vector and initial object affinity vector with at least one additional dimension and produce predicted affinity values based on the modified initial subject affinity vector and modified initial object affinity vector, and to determine a cost function for the modified initial subject affinity vector and modified initial object affinity vector, based on the predicted affinity values; and wherein the computer server is further configured to iteratively add additional dimensions to the generated subject affinity vectors and to the generated object affinity vectors to produce modified subject affinity vectors and a modified object affinity vectors and new predicted affinity values, and to determine a new cost function for the modified subject affinity vectors and the modified object affinity vectors, based on the new predicted affinity values, such that the computer server halts the modifying and uses the last modified subject affinity vector and last modified object affinity vectors as the subject affinity vector and the object affinity vector for the object recommendations, in response to a value of the cost function being equal to or less than a predetermined value of the cost function.
-
Specification