Method for estimating user interests
First Claim
1. A computer-implemented method for estimating user interests, the method executable by a computing device in communication with an output device, the method comprising:
- determining, by the computing device, a first input vector corresponding to a first user event and a second input vector corresponding to a separate second user event;
configuring a first vector-mapping module to map given input vectors to output vectors in a first multidimensional space such that a respective distance separating the output vectors from each other is correlated to a difference between a user event context associated with the first input vector having been mapped to its output vector and a user event context associated with the second input vector having been mapped to its output vector, the configuring the first vector-mapping module including configuring a first neural network and a second neural network by;
connecting the first neural network and the second neural network in a coupled Siamese neural network arrangement, andtraining the first and second neural networks to minimize a cross-modal loss between the first neural network and the second neural network;
mapping, using the first vector-mapping module of the computing device, the first input vector to a first output vector in the first multidimensional space and the separate second input vector to a second output vector in the first multidimensional space,the first vector-mapping module being enabled to map both the first output vector and the second output vector to the first multidimensional space even if the first user event and the second user event are of different types;
determining, by the computing device, a third input vector based at least in part on the first output vector and the separate second output vector;
configuring a second vector-mapping module to map input vectors to output vectors in a second multidimensional space such that a respective distance separating the output vectors from each other is correlated to a difference between a user event context associated with one input vector, the one input vector being a respective output vector of the first vector-mapping module and a user event context associated with another input, the other input vector being another respective output vector of the first vector-mapping module, the configuring the second vector-mapping module including configuring a third neural network;
mapping, using the second vector-mapping module of the computing device, the third input vector to a third output vector in the second multidimensional space;
determining, by the computing device, a message to be provided to a user based at least in part on an analysis of at least one of the first output vector and the third output vector; and
causing, by the computing device, the output device to provide the message to the user.
4 Assignments
0 Petitions
Accused Products
Abstract
Computer-implemented method for estimating user interests, executable by a computing device in communication with an output device, comprising: determining a first input vector corresponding to a first user event and a second input vector corresponding to a second user event; mapping first input vector to a first output vector and second input vector to a second 5 output vector in a first multidimensional space using a first vector-mapping module; determining a third input vector based on first output vector and second output vector; mapping third input vector to a third output vector in a second multidimensional space using a second vector-mapping module; determining a message to be provided to a user based on an analysis of at least one of first output vector and third output vector; and causing output 10 device to provide message to user. Also non-transitory computer-readable medium storing program instructions for carrying out the method.
-
Citations
25 Claims
-
1. A computer-implemented method for estimating user interests, the method executable by a computing device in communication with an output device, the method comprising:
-
determining, by the computing device, a first input vector corresponding to a first user event and a second input vector corresponding to a separate second user event; configuring a first vector-mapping module to map given input vectors to output vectors in a first multidimensional space such that a respective distance separating the output vectors from each other is correlated to a difference between a user event context associated with the first input vector having been mapped to its output vector and a user event context associated with the second input vector having been mapped to its output vector, the configuring the first vector-mapping module including configuring a first neural network and a second neural network by; connecting the first neural network and the second neural network in a coupled Siamese neural network arrangement, and training the first and second neural networks to minimize a cross-modal loss between the first neural network and the second neural network; mapping, using the first vector-mapping module of the computing device, the first input vector to a first output vector in the first multidimensional space and the separate second input vector to a second output vector in the first multidimensional space, the first vector-mapping module being enabled to map both the first output vector and the second output vector to the first multidimensional space even if the first user event and the second user event are of different types; determining, by the computing device, a third input vector based at least in part on the first output vector and the separate second output vector; configuring a second vector-mapping module to map input vectors to output vectors in a second multidimensional space such that a respective distance separating the output vectors from each other is correlated to a difference between a user event context associated with one input vector, the one input vector being a respective output vector of the first vector-mapping module and a user event context associated with another input, the other input vector being another respective output vector of the first vector-mapping module, the configuring the second vector-mapping module including configuring a third neural network; mapping, using the second vector-mapping module of the computing device, the third input vector to a third output vector in the second multidimensional space; determining, by the computing device, a message to be provided to a user based at least in part on an analysis of at least one of the first output vector and the third output vector; and causing, by the computing device, the output device to provide the message to the user. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24)
-
-
25. A non-transitory computer-readable medium storing program instructions for estimating user interests, the program instructions being executable by a computing device in communication with an output device to effect:
-
determination, by the computing device, of a first input vector corresponding to a first user event and a second input vector corresponding to a separate second user event; configuring a first vector-mapping module to map given input vectors to output vectors in a first multidimensional space such that a respective distance separating the output vectors from each other is correlated to a difference between a user event context associated with the first input vector having been mapped to its output vector and a user event context associated with the second input vector having been mapped to its output vector, the configuring the first vector-mapping module including configuring a first neural network and a second neural network by; connecting the first neural network and the second neural network in a coupled Siamese neural network arrangement, and training the first and second neural networks to minimize a cross-modal loss between the first neural network and the second neural network; mapping, using a first vector-mapping module of the computing device, of the first input vector to a first output vector in a first multidimensional space and the separate second input vector to a second output vector in the first multidimensional space, the first vector-mapping module being enabled to map both the first output vector and the second output vector to the first multidimensional space even if the first user event and the second user event are of different types; determination, by the computing device, of a third input vector based at least in part on the first output vector and the separate second output vector; configuring a second vector-mapping module to map input vectors to output vectors in a second multidimensional space such that a respective distance separating the output vectors from each other is correlated to a difference between a user event context associated with one input vector, the one input vector being a respective output vector of the first vector-mapping module and a user event context associated with another input, the other input vector being another respective output vector of the first vector-mapping module, the configuring the second vector-mapping module including configuring a third neural network; mapping, using the second vector-mapping module of the computing device, of the third input vector to a third output vector in the second multidimensional space; determination, by the computing device, of a message to be provided to a user based at least in part on an analysis of at least one of the first output vector and the third output vector; and causing, by the computing device, the output device to provide the message to the user.
-
Specification