Inferential People Recommender Method and System

Abstract

An inferential people recommender method and system infers the preferences of users from usage behaviors and generates recommendations of people based upon the matching of the inferred preferences. Natural language-based explanations that provide the reasoning that is applied in making the recommendations are also provided. The inferences of preferences may be determined by inference weightings that are in accordance with behavioral-based priority rules. The inferences may be further based on factors such as proximity, and duration of proximity, of people to physical objects, including other people.

What is claimed is: 
     
         1 . A method, comprising:
 inferring automatically a preference of a first person who is a first user of a computer-implemented system from a first plurality of usage behaviors;   inferring automatically a preference of a second person who is a second user of the computer-implemented system from a second plurality of usage behaviors;   matching automatically the preference of the first person and the preference of the second person;   generating automatically a recommendation based upon the matching, wherein the recommendation comprises a reference to the second person;   delivering automatically the recommendation to the first user; and   delivering automatically an explanation for the recommendation to the first user, wherein the explanation is in a natural language format and comprises reasoning for the delivery of the recommendation to the first user.   
     
     
         2 . The method of  claim 1 , further comprising:
 inferring automatically the preference of the first user of the computer-implemented system from the first plurality of usage behaviors, wherein the inferring of the preference that is based upon the application of a plurality of inference weightings that are determined in accordance with usage behavior priority rules that are applied to the first plurality of usage behaviors.   
     
     
         3 . The method of  claim 2 , further comprising:
 inferring the preference that is based upon the application of the plurality of inference weightings that are determined in accordance with the usage behavior priority rules that are applied to the first plurality of usage behaviors, wherein at least one of the inference weightings is based upon a user's duration of attention toward a computer-implemented object.   
     
     
         4 . The method of  claim 2 , further comprising:
 inferring the preference that is based upon the application of the plurality of inference weightings that are determined in accordance with the usage behavior priority rules that are applied to the first plurality of usage behaviors, wherein at least one of the inference weightings is based upon a user's duration of being proximal to a physical object.   
     
     
         5 . The method of  claim 1 , further comprising:
 inferring automatically the preference of the first user of the computer-implemented system, wherein the preference is further inferred from an automatic analysis of computer-implemented content.   
     
     
         6 . The method of  claim 1 , further comprising:
 generating automatically the recommendation based upon the matching, wherein the recommendation further comprises self-profiling information provided by the second person.   
     
     
         7 . The method of  claim 1 , further comprising:
 delivering automatically the explanation for the recommendation to the first user, wherein the explanation further comprises a reference to at least one usage behavior of the second plurality of usage behaviors.   
     
     
         8 . A computer-implemented system comprising one or more processors configured to:
 infer a preference of a first person who is a first user of the computer-implemented system from a first plurality of usage behaviors;   infer a preference of a second person who is a second user of the computer-implemented system from a second plurality of usage behaviors;   match the preference of the first person and the preference of the second person;   generate a recommendation based upon the matching, wherein the recommendation comprises a reference to the second person;   deliver the recommendation to the first user; and   deliver an explanation for the recommendation to the first user, wherein the explanation is in a natural language format and comprises reasoning for the delivery of the recommendation to the first user.   
     
     
         9 . The system of  claim 8  comprising the one or more processors, further configured to:
 infer the preference of the first user of the computer-implemented system from the first plurality of usage behaviors, wherein the inference of the preference is based upon the application of a plurality of inference weightings that are determined in accordance with usage behavior priority rules that are applied to the first plurality of usage behaviors. 
 
     
     
         10 . The system of  claim 9  comprising the one or more processors, further configured to:
 infer the preference that is based upon the application of the plurality of inference weightings that are determined in accordance with the usage behavior priority rules that are applied to the first plurality of usage behaviors, wherein at least one of the inference weightings is based upon a user's duration of attention toward a computer-implemented object that represents a person. 
 
     
     
         11 . The system of  claim 9  comprising the one or more processors, further configured to:
 infer the preference that is based upon the application of the plurality of inference weightings that are determined in accordance with the usage behavior priority rules that are applied to the first plurality of usage behaviors, wherein at least one of the inference weightings is based upon a user's duration of proximity to a physical object. 
 
     
     
         12 . The system of  claim 8  comprising the one or more processors, further configured to:
 infer the preference of the first user of the computer-implemented system, wherein the preference is further inferred from an automatic analysis of computer-implemented content. 
 
     
     
         13 . The system of  claim 12  comprising the one or more processors, further configured to:
 infer the preference of the first user of the computer-implemented system, wherein the preference is further inferred from the automatic analysis, wherein the automatic analysis is performed through application of a computer-implemented neural network. 
 
     
     
         14 . The system of  claim 13  comprising the one or more processors, further configured to:
 deliver the explanation for the recommendation to the first user, wherein the explanation further comprises a reference to at least one usage behavior of the second plurality of usage behaviors. 
 
     
     
         15 . A computer-implemented system comprising one or more processors configured to:
 infer a preference of a first person who is a user of a first mobile device from a first plurality of usage behaviors;   infer a preference of a second person who is a user of a second mobile device from a second plurality of usage behaviors;   match the preference of the first person and the preference of the second person;   generate a recommendation based upon the matching, wherein the recommendation comprises a reference to the second person;   deliver the recommendation to the user of the first mobile device; and   deliver an explanation for the recommendation to the user of the first mobile device, wherein the explanation is in a natural language format and comprises reasoning for the delivery of the recommendation to the first user.   
     
     
         16 . The system of  claim 15  comprising the one or more processors, further configured to:
 infer the preference of the user of the first mobile device from the first plurality of usage behaviors, wherein the inference of the preference is based upon the application of a plurality of inference weightings that are determined in accordance with usage behavior priority rules that are applied to the first plurality of usage behaviors. 
 
     
     
         17 . The system of  claim 16  comprising the one or more processors, further configured to:
 infer the preference that is based upon the application of the plurality of inference weightings that are determined in accordance with the usage behavior priority rules that are applied to the first plurality of usage behaviors, wherein at least one of the inference weightings is based upon a proximity of the user of the first mobile device to a physical object. 
 
     
     
         18 . The system of  claim 17  comprising the one or more processors, further configured to:
 infer the preference that is based upon the application of the plurality of inference weightings that are determined in accordance with the usage behavior priority rules that are applied to the first plurality of usage behaviors, wherein at least one of the inference weightings is based upon the proximity of the user of the first mobile device to the physical object, wherein the physical object comprises the second person. 
 
     
     
         19 . The system of  claim 15  comprising the one or more processors, further configured to:
 infer the preference of the first user of the computer-implemented system, wherein the preference is further inferred from an automatic analysis of pictorial-based information, wherein the automatic analysis is performed through application of a computer-implemented neural network. 
 
     
     
         20 . The system of  claim 15  comprising the one or more processors, further configured to:
 deliver the explanation for the recommendation to the user of the first mobile device, wherein the explanation further comprises a reference to a geographic location of the user of the second mobile device.