USER-CONTROLLABLE LEARNING OF POLICIES
First Claim
1. A computer implemented method for updating a policy that is enforced by a computer program, the method comprising:
- communicating, by a computer system to at least one user, data regarding one or more decisions made by the program over a period of time according to a policy, wherein each decision is made on the particular policy in force at the time the decision is made, and wherein policy data for the policy is stored in a machine readable format in a memory coupled to the computer system, wherein the computer system comprises at least one processor coupled to a memory;
storing in a machine readable format user feedback data indicative of feedback by the at least one user regarding the one or more decisions made by the program according to the policy;
identifying and ranking, by the computer system, one or more potential variations to the policy based on a score of an objective function, wherein the objective function comprises a plurality of function elements, wherein the plurality of function elements comprises a user satisfaction function element and at least one other function element that measures an extent to which the one or more potential variations are likely to be understandable by the at least one user;
communicating, by the computer system to the at least one user, one or more suggested modifications to the policy based on the ranking of the one or more potential variations to the policy, wherein the one or more suggested modifications can assist the one or more users refine the policy; and
modifying, by the computer system, the policy data stored in the memory based on one or more selections by the at least one user in response to the communication of the one or more suggested modifications.
10 Assignments
0 Petitions
Accused Products
Abstract
Various embodiments are directed to a computer implemented method for updating a policy that is enforced by a computer program. In one embodiment, a computer communicates, to a user, data regarding one or more decisions made by the program over a period of time according to a policy. Each decision is made on the particular policy in force at the time the decision is made. Policy data for the policy is stored in a machine readable format. The user feedback data indicative of feedback by the user regarding the one or more decisions is stored. The computer identifies and ranks one or more potential variations to the policy based on a score of an objective function for each potential variation. The computer communicates, to the user, one or more suggested modifications based on the ranking of the one or more potential variations to the policy. The computer modifies the policy data based on one or more selections by the user in response to the transmission of the one or more suggested modifications.
148 Citations
93 Claims
-
1. A computer implemented method for updating a policy that is enforced by a computer program, the method comprising:
-
communicating, by a computer system to at least one user, data regarding one or more decisions made by the program over a period of time according to a policy, wherein each decision is made on the particular policy in force at the time the decision is made, and wherein policy data for the policy is stored in a machine readable format in a memory coupled to the computer system, wherein the computer system comprises at least one processor coupled to a memory; storing in a machine readable format user feedback data indicative of feedback by the at least one user regarding the one or more decisions made by the program according to the policy; identifying and ranking, by the computer system, one or more potential variations to the policy based on a score of an objective function, wherein the objective function comprises a plurality of function elements, wherein the plurality of function elements comprises a user satisfaction function element and at least one other function element that measures an extent to which the one or more potential variations are likely to be understandable by the at least one user; communicating, by the computer system to the at least one user, one or more suggested modifications to the policy based on the ranking of the one or more potential variations to the policy, wherein the one or more suggested modifications can assist the one or more users refine the policy; and modifying, by the computer system, the policy data stored in the memory based on one or more selections by the at least one user in response to the communication of the one or more suggested modifications. - 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, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46)
-
-
47. A computer system for updating a policy that is enforced by a computer program, the computer system comprising:
-
at least one processor and a memory coupled to the at least one processor, wherein the at least one processor is to; communicate to at least one user data regarding one or more decisions made by the program over a period of time according to a policy, wherein each decision was made on the particular policy in force at the time the decision was made, and wherein policy data for the policy is stored in a machine readable format in a memory coupled to the computer system; store in a machine readable format user feedback data indicative of feedback by the at least one user regarding the one or more decisions made by the program according to the policy; identify and rank one or more potential variations to the policy based on a score of an objective function, wherein the objective function comprises a plurality of function elements, and wherein the plurality of function elements comprises a user satisfaction function element and at least one other function element that measures an extent to which the one or more potential variations are likely to be understandable by the at least one user; communicate to the at least one user, one or more suggested modifications to the policy based on the ranking of the one or more potential variations to the policy, wherein the one or more suggested modifications can assist the at least one user refine the policy; and modify the policy data stored in the memory based on one or more selections by the at least one user in response to the communication of the one or more suggested modifications.
-
-
48-92. -92. (canceled)
-
93. An article comprising a computer-readable storage medium comprising executable computer program instructions that when executed enable a computing system to:
-
communicate to at least one user data regarding one or more decisions made by the program over a period of time according to a policy, wherein each decision was made on the particular policy in force at the time the decision was made, and wherein policy data for the policy is stored in a machine readable format in a memory coupled to the computer system; store in a machine readable format user feedback data indicative of feedback by the at least one user regarding the one or more decisions made by the program according to the policy; identify and rank one or more potential variations to the policy based on a score of an objective function, wherein the objective function comprises a plurality of function elements, and wherein the plurality of function elements comprises a user satisfaction function element and at least one other function element that measures an extent to which the one or more potential variations are likely to be understandable by the at least one user; communicate to the at least one user, one or more suggested modifications to the policy based on the ranking of the one or more potential variations to the policy, wherein the one or more suggested modifications can assist the at least one user refine the policy; and modify the policy data stored in the memory based on one or more selections by the at least one user in response to the communication of the one or more suggested modifications.
-
Specification