Method and system for matching entities in an auction

• US 10,237,420 B1
• Filed: 12/28/2017
• Issued: 03/19/2019
• Est. Priority Date: 03/07/2003
• Status: Expired due to Term

• Current Assignee: Patent Armory Inc. (Jeffrey M. Gross)
• Original Assignee: Steven Mark Hoffberg, Wai Wu
• Inventors: Wu, Wai, Hoffberg, Steven M.
• Primary Examiner(s): Nguyen, Khai N.
• Application Number:
  US15/856,729
• Time in Patent Office: 446 Days
• Field of Search: 37921001, 37926501, 37926502, 37926503, 37926504, 37926505, 37926506, 37926507, 37926508, 37926509, 3792651, 37926511, 37926512, 37926513, 37926514, 37926601, 37926602, 37926603, 37926604, 37926605, 37926606, 37926607, 37926608, 37926609, 3792661
• CPC Class Codes: H04M 15/8061 Selecting least cost route depending on origin or type of service
• CPC Class Codes: H04M 3/42068 where the identifier is used to access a profile
• CPC Class Codes: H04M 3/523 with call distribution or queueing
• CPC Class Codes: H04M 3/5233 Operator skill based call distribution
• CPC Class Codes: H04M 3/5238 with waiting time or load prediction arrangements

Abstract

A method for matching a first entity with at least one second entity selected from a plurality of second entities, comprising defining a plurality of multivalued scalar data representing inferential targeting parameters for the first entity and a plurality of multivalued scalar data of each of the plurality of second entities, representing respective characteristic parameters for each respective second entity; and performing an automated optimization with respect to an economic surplus of a respective match of the first entity with the at least one of the plurality of second entities, and an opportunity cost of the unavailability of the at least one of the plurality of second entities for matching with an alternate first entity.

Claims

1. A method of pairing requests, comprising:

estimating at least one content-specific or requestor-specific characteristic associated with a request;

determining a set of available partners, each having at least one respective partner characteristic;

evaluating, with at least one automated processor, a plurality of pairings of the request with a plurality of different available partners, according to an evaluator for valuing pairings of the request with respective available partners, comprising at least one of a probabilistic function, a hierarchical Markov model, Bayesian logic, and a neural network, dependent on the at least one content-specific or requestor-specific characteristic, and the at least one respective partner characteristic; and

generating a control signal, by the at least one automated processor, selectively dependent on the evaluating.

2. The method according to claim 1, wherein the evaluator comprises an economic function further dependent on a pairing cost or benefit of the requestor with respective partners.

3. The method according to claim 2, wherein the evaluator further comprises an auction, further comprising receiving bids from respective partners.

4. The method according to claim 1, wherein the evaluator comprises a multivariate function of at least two content-specific or requestor-specific characteristics and at least two respective partner characteristics.

5. The method according to claim 1, wherein the evaluator function comprises a clustering algorithm.

6. The method according to claim 5, wherein the clustering algorithm comprises a multidimensional clustering algorithm for clustering data in a space having at least 3 dimensions.

7. The method according to claim 1, wherein the evaluator comprises a hierarchical Markov model.

8. The method according to claim 1, wherein the evaluator comprises a Bayesian network.

9. The method according to claim 1, wherein the evaluator comprises a neural network.

10. The method according to claim 1, wherein the evaluator is dependent on a shape of at least one probability distribution.

11. The method according to claim 1, wherein the at least one content-specific or requestor-specific characteristic comprises at least one of a customer satisfaction and a user profile.

12. The method according to claim 1, wherein the at least one content-specific or requestor-specific characteristic is derived from a past history comprising a plurality of transactions.

13. The method according to claim 1, wherein the control signal changes an availability of at least one partner, further comprising updating a memory storing partner availability information.

14. The method according to claim 1, wherein the request comprises a request for communication, further comprising establishing the requested communication by controlling a communication router selectively dependent on the control signal.

15. The method according to claim 1, wherein the request comprises a voice call, further comprising connecting the voice call with a partner selectively dependent on the control signal.

16. The method according to claim 1, wherein the request is related to a transaction, having an economic outcome, wherein the evaluator is dependent on a statistical function optimized to indicate a pairing having a highest probabilistic economic outcome.

17. The method according to claim 1, further comprising:

receiving a series of requests;

generating a series of control signals which concurrently alter availability of a plurality of partners; and

dynamically updating availability of partners over time;

wherein the evaluator maximizes an aggregate value of the plurality of available partners.

18. The method according to claim 1, wherein the evaluator performs at least one of a combinatorial analysis and an opportunity-cost dependent analysis of a plurality of alternate pairings.

19. A system for pairing requests, comprising:

at least one port, configured to receive a request and information identifying a content or requestor-related characteristic associated with the request;

at least one memory, configured to store availability information for a plurality of partners, and at least one respective partner characteristic associated with each respective available partner; and

at least one automated processor, configured to generate a control signal, selectively dependent on an evaluation of each of a plurality of pairings of the request with a plurality of different available partners, dependent on at least one probabilistic function of the at least one content or requestor-related characteristic associated with the request, and the at least one respective partner characteristic.

20. A method of processing requests, comprising:

estimating at least one content-specific or requestor-specific characteristic associated with each received request;

determining availability of a plurality of alternate target resources, each respective target resource having at least one respective target characteristic;

evaluating, with at least one automated processor, a plurality of alternate allocations of the respective received request with different available targets, according to a ranking dependent on a probabilistic predictive multivariate evaluator, based on the at least one content-specific or requestor-specific characteristic, and the respective target characteristics of the plurality of alternate target resources; and

generating a control signal, by the at least one automated processor, selectively dependent on the evaluating, to control the allocations of the respective received request with the different available targets.