System and method for determining contingent relevance

• US 9,818,136 B1
• Filed: 02/03/2004
• Issued: 11/14/2017
• Est. Priority Date: 02/05/2003
• Status: Active Grant

• Current Assignee: Steven Mark Hoffberg
• Original Assignee: Steven Mark Hoffberg
• Inventors: Hoffberg, Steven M.
• Primary Examiner(s): Ouellette, Jonathan
• Application Number:
  US10/771,182
• Time in Patent Office: 5,033 Days
• Field of Search: 705 1, 705 10, 705 26, 705 27, 705 37, 705 80, 705347, 705 11-912
• CPC Class Codes: G06Q 30/0207 Discounts or incentives, e.g. coupons or rebates
• CPC Class Codes: G06Q 30/0282 Rating or review of business operators or products
• CPC Class Codes: G06Q 30/08 Auctions
• CPC Class Codes: G07F 17/32 for games, toys, sports, or amusements
• CPC Class Codes: G07F 17/323 wherein the player is informed, e.g. advertisements, odds, instructions
• CPC Class Codes: G07F 17/3237 about the players, e.g. profiling, responsible gaming, strategy/behavior of players, location of players

Abstract

A system and method providing for communication and resolution of utility functions between participants, wherein the utility function is evaluated based on local information at the recipient to determine a cost value thereof. A user interface having express representation of both information elements, and associated reliability of the information. An automated system for optimally conveying information based on relevance and reliability.

Claims

1. A method, comprising:

receiving at least one sensor signal from at least one condition sensor;

analyzing the at least one sensor signal with an automated processor, to determine;

a plurality of possible events represented in the at least one sensor signal, and

a respective quantitative probability of at least one respective contingency associated with at least one respective possible event of the plurality of possible events;

defining an interest or context of at least one user;

determining, by the automated processor;

a contingent relevance to the defined interest or context of the at least one user, and

an associated statistical distribution of the contingent relevance to the defined interest or context of the at least one user, of the determined plurality of possible events,

based on at least the respective quantitative probability of the at least one respective contingency associated with the at least one respective possible event of the plurality of possible events;

ranking the determined plurality of possible events, with the automated processor, dependent on at least both;

the determined contingent relevance to the defined interest or context of the at least one user, and

the associated statistical distribution of the contingent relevance to the defined interest or context of the at least one user,

wherein the ranking differs from a ranking dependent on only one of the determined contingent relevance to the defined interest or context of the at least one user, and the associated statistical distribution of the contingent relevance to the defined interest or context of the at least one user; and

producing an output, by the automated processor, of a control signal, for controlling a system interacting with a condition sensed by the at least one condition sensor, selectively dependent on the ranking of the determined plurality of possible events, in advance of the respective at least one respective contingency associated with the at least one respective possible event of the plurality of possible events.

2. The method according to claim 1, wherein said associated statistical distribution describes a quantitative probability of existence of a particular possible event, and said relevance to the defined interest or context of the user comprises a value function associated with that particular possible event, wherein said ranking comprises quantitative analysis of probability-weighted benefits from each particular possible event to an overall utility function for the user.

3. The method according to claim 1, wherein said ranking comprises a combinatorial analysis of a set of competing potential ranked groups of the determined plurality of possible events, conducted by the automated processor, and the controlled system is a vehicular system.

4. An apparatus comprising:

a memory configured to store at least one of user relevance and user context information for a respective user;

an input configured to receive at least one sensor signal from at least one condition sensor;

an automatically-implemented artificial intelligence configured to analyze the at least one sensor signal, to determine a plurality of possible events represented in the at least one sensor signal and a respective quantitative probability of at least one respective contingency associated with at least one possible event of the plurality of possible events;

an automated processor configured to;

determine, with respect to the plurality of possible events and based on the user relevance information, the determined plurality of possible events represented in the at least one sensor signal, and the respective quantitative probability of the at least one respective contingency associated with the at least one possible event of the plurality of possible events, a contingent relevance of the plurality of possible events to the respective user and an associated statistical distribution of the contingent relevance of the plurality of possible events to the respective user; and

rank the determined plurality of possible events selectively dependent on at least both the determined contingent relevance of the plurality of possible events to the user, and the associated statistical distribution of the contingent relevance of the plurality of possible events to the respective user, to produce a ranked set, wherein a rank of respective possible events within the ranked set is dependent on both the contingent relevance of the plurality of possible events to the respective user and the associated statistical distribution of the contingent relevance of the plurality of possible events to the respective user; and

an output configured to convey a control signal for a controlled system which interacts with a condition sensed by the at least one condition sensor, selectively dependent on the ranked set, in advance of the respective at least one respective contingency associated with at least one possible event of the plurality of possible events.

5. The apparatus according to claim 4, wherein said associated statistical distribution of the contingent relevance of the plurality of possible events to the respective user describes a probability of existence of an associated event, and said contingent relevance of the plurality of possible events to the respective user comprises a value function associated with that associated event if it exists, wherein said rank of respective ones of the determined plurality of possible events is dependent on at least a quantitative analysis of probability-weighted benefits from each associated event to an overall utility function for the respective user.

6. The apparatus according to claim 4, wherein said automated processor is further configured to perform a combinatorial analysis of a plurality of competing ranked sets to select an optimal ranked set, and the controlled system comprises a vehicular system.

7. A computer readable medium storing nontransitory instructions therein for controlling a programmable processor to perform operations comprising:

receiving at least one sensor signal from at least one condition sensor;

storing user relevance information for a respective user;

analyzing the at least one sensor signal and the stored user relevance information, to determine possible events represented in the at least one sensor signal, and a respective quantitative probability of at least one respective contingency associated with at least one of the plurality of possible events;

determining, based on at least the determined plurality of possible events and the respective quantitative probability of the at least one respective contingency associated with the at least one of the plurality of possible events, a contingent relevance of the plurality of possible events to the respective user and an associated statistical distribution of the contingent relevance of the determined plurality of possible events to the respective user;

ranking the plurality of possible events in advance of the respective at least one respective contingency associated with the at least one of the plurality of possible events, dependent on both the contingent relevance of the plurality of possible events to the respective user and the associated statistical distribution of the contingent relevance of the determined plurality of possible events to the respective user; and

presenting an a control signal output, for controlling a system which interacts with a condition sensed by the at least one condition sensor, in advance of the respective at least one respective contingency associated with the plurality of possible events, selectively dependent on the ranked plurality of possible events.

8. The computer readable medium according to claim 7, wherein said associated statistical distribution of the contingent relevance of the determined plurality of possible events to the respective user describes a probability of existence of an associated event, and said contingent relevance of the determined plurality of possible events to the respective user comprises a quantitative value function associated with that event if it exists, wherein said ranking comprises a quantitative analysis of probability-weighted benefits from each of the plurality of possible events to an overall utility function for the respective user.

9. The computer readable medium according to claim 7, wherein said ranking comprises performing a combinatorial analysis of competing sets of rankings, and the system comprises a vehicular system.

10. A method, comprising:

receiving information from at least one sensor, corresponding to a plurality of possible events and contingent predicates to the plurality of possible events;

defining a contingent relevance of the information from the at least one sensor to a user, based at least on a user-relevance model stored in a memory and a respective quantitative statistical distribution of the contingent relevance of the information from the at least one sensor to the user, with respect to at least one independent variable, distinct from the contingent relevance of the information from the at least one sensor to the user, comprising at least time;

automatically ranking, by an automated computing device, at least a portion of the plurality of possible events, based on at least the contingent relevance of the information from the at least one sensor to the user, the respective statistical distribution of the contingent relevance of the information from the at least one sensor to the user, and the at least one independent variable comprising at least time; and

producing an output in accordance with said ranking, before the contingent predicates are realized and truncated with respect to possible events determined to be inconsistent with the received information from the at least one sensor, representing a control signal for a controlled system which has a possible interaction with at least one possible event,

wherein the automatically ranking based on at least the contingent relevance of the information from the at least one sensor to the user, the respective statistical distribution of the contingent relevance of the information from the at least one sensor to the user, and the at least one independent variable comprising at least time, differs from a ranking solely based on a non-contingent relevance.

11. The method according to claim 10, wherein the at least one independent variable comprises distance, and the controlled system comprises a vehicular system.

12. A method, comprising:

receiving information from at least one sensor, corresponding to quantitative statistical predicates to a plurality of possible future events, contingent on the respective quantitative statistical predicates, and information representing subsequent occurrence of the possible events;

retrieving information from a memory, associated with user relevance of the possible future events, comprising at least one subjective criterion for a respective user;

automatically ranking at least a portion of the plurality of possible future events, based on at least both a time-independent contingent relevance of a respective possible future event to the respective user with respect to the at least the at least one subjective criterion, contingent on the respective quantitative statistical predicates, and a respective statistical distribution of the contingent relevance of the plurality of possible future events to the user over time; and

producing an output, by an automated processor apparatus, representing a control system for a controlled system, based on said automatically ranking, of the at least a portion of the plurality of future events, adapted to alter a probability of at least one future event by an action of the controlled system.

13. A method, comprising:

receiving information corresponding to a quantitative statistical likelihood of occurrence of a plurality of future events, a likely time of occurrence of the plurality of future events, and a contingent user-relevance of the plurality of future events;

automatically ranking at least a portion of at least one of the plurality of future events, by an automated processing device, based on at least the contingent user-relevance of the future events, the quantitative statistical likelihood of occurrence of the plurality of future events, the likely time of occurrence of the plurality of future events, and a statistical change in user-relevance of respective future events of the plurality of future events over time; and

generating at least one output representing a control signal for a controlled system, for altering at least one future event based on a response of the controlled system to the control signal.

14. An method for controlling a system, comprising:

receiving a sensor signal representing information relating to a plurality of possible future events;

storing a time-dependent relevance function of the plurality of possible future events associated with a user, wherein a current relevance of at least one possible future event varies with respect to a time of occurrence of the possible future event;

automatically ranking, using a computer, the plurality of possible future events, based on at least the respective time-dependent relevance function of respective possible future events, and a probability distribution of a quantitative likelihood of occurrence of the possible future events;

producing a control signal dependent on said ranking, adapted to control a system interacting with at least one of the plurality of possible future events; and

controlling the system based on the produced control signal to alter at least one possible future event.

15. A method of controlling a system, comprising:

defining a user relevance of a plurality of subjects in a stream of data from a condition sensor, according to a respective subject content and a respective subject context, wherein at least one respective subject context has a stochastic distribution with respect to an independent variable distinct from the respective subject content;

automatically determining a relative user-dependent ranking of the plurality of subjects over at least a portion of a range of the independent variable, based on at least the defined user relevance and a state of the respective subject context, with respect to the stochastic distribution with respect to the independent variable distinct from the respective subject content, over the at least a portion of the range of the independent variable;

at least one of outputting or storing, by a computer, a ranked set comprising at least a portion of the relative user-dependent ranked plurality of subjects not inconsistent with the stream of data prior to said at least one of outputting or storing; and

automatically producing at least one control signal for a controlled system, the controlled system being adapted for modifying future states of the stream of data.

16. A method for controlling a system, according to a ranking of objects having a quantitative relevance to a user whose statistical magnitude varies over a range of a time domain process demonstrating a Markov property, contingent on a plurality of future states having associated conditional probabilities, comprising:

receiving information sufficient to identify the objects from at least one condition sensor;

receiving information describing a quantitative distribution of relevance of the objects to the user over the range of the time domain process;

ranking, by an automated processor, at least a portion of the objects, based on at least the respective quantitative distribution of relevance of respective objects to the user, and a stochastic distribution of the plurality of future states having the associated conditional probabilities, representing at least a portion of the range of the time domain process; and

producing a control signal for a controlled system, the controlled system being configured to respond to the control signal as a feedback signal, and to produce an action to modify a statistical probability of at least one respective future state.