Predictive bounding of combinatorial optimizations that are based on data sets acquired post-prediction through high-latency, heterogenous interfaces

Patent metadata

Field	Value
Publication number	US-11328313-B2
Application number	US-202016870172-A
Country	US
Kind code	B2
Filing date	May 8, 2020
Priority date	May 8, 2019
Publication date	May 10, 2022
Grant date	May 10, 2022

Abstract

Disclosed are techniques for achieving high-volumes of low-latency price plus discount transactions that are dynamically created through transaction optimization. Aspects may include combining and optimizing different types of discounts into a single discount and dynamically controlling the single discount through a single user-facing transaction. A computer model may determine one or more transactions and sets of discounts applicable to those transactions upon which the user-facing transaction and single discount are based. Associations between discounts and items may be determined for at least some different types of discounts to apply a set of discounts including at least two different discounts of different discount types to a desired transaction for one or more items. Aspects may include determining a discount transaction comprising a plurality of constituent transactions that form a single user-facing transaction.

First claim

What is claimed is:

1. A tangible, non-transitory, machine-readable medium storing computing program instructions that when executed by one or more processors cause the processors to perform operations comprising:
obtaining, with a computer system, a set of item records corresponding to a plurality of items, an item record in the set indicating one or more merchants offering a corresponding item to users;
obtaining, with the computer system, a set of discount records corresponding to a plurality of discounts, a discount record in the set indicating one or more discounts applicable to a merchant or an item, each of the one or more discounts having a respective discount type;
determining, with the computer system, associations between discounts and items based on the set of items records and the set of discount records, at least some of the items associated with at least two discounts having different discount types;
training, with the computer system, based at least in part on the set of item records and the set of discount records, a computer model to determine a discount transaction with a given merchant by estimating the discount transaction provides a lower price for an item of a desired transaction responsive to an application of an optimal set of discounts with the given merchant relative to prices corresponding to a plurality of other merchants and a plurality of other sets of discounts applicable to respective ones of the other merchants;
after training, obtaining, with the computer system, a desired transaction and a geolocation of a user, wherein the desired transaction is indicative of one or more items a user desires to purchase;
obtaining, with the computer system, one or more item records corresponding to the one or more items based on the desired transaction and data indicative of the associations of the item records with discounts;
determining, with the computer system, with the trained computer model, based on the user record, the item records, and the associated discounts, at least one discount transaction comprising an estimated discounted price and a price for the one or more items responsive to an optimal set of discounts determined for the discount transaction with a given one of the merchants; and
in response to receiving a confirmation for the discount transaction, automatically, with the computer system, effectuating a transaction with the given one of the merchants for one or more of the items provided by the merchant based on the discount transaction and the user record by applying each of two or more discounts in the optimal set of discounts.

2. The medium of claim 1 , wherein:
applying at least one of the discounts in the set comprises provisioning a first discount of a first discount type that satisfies a transaction price with the given one of the merchants, the transaction price with the given one of the merchants being different from and greater than the estimated discount price.

3. The medium of claim 2 , wherein the transaction price is unaffected by at least one of the other applied discounts in the set.

4. The medium of claim 1 , wherein training, based at least in part on the first and second sets of records, comprises:
training a computer model configured to determine a discount transaction comprising of a set of constituent transactions, a constituent transaction estimated to provide a given price for one or more items with a given merchant responsive to an application of a set of discounts with the given merchant out of a plurality of other merchants and a plurality of other sets of discounts applicable to respective ones of the other merchants, the set of constituent transactions comprising two or more constituent transactions estimated to provide a greater discount on a plurality of items across at least two merchants based on a respective optimal set of discounts for each merchant than for any single transaction with a single merchant based on an optimal set of discounts for the single merchant.

5. The medium of claim 4 , wherein automatically effectuating the discount transaction with the given one of the merchants comprises:
processing a given one of the two or more constituent transactions by:
provisioning one or more stored value cards redeemable at a respective merchant specified by the given one of the two or more constituent transactions with a collective stored value corresponding to or exceeding a transaction price with the merchant for corresponding ones of the items; and
executing a sequence of steps to apply each of two or more discounts in the optimal set of discounts determined for the given one of the constituent transactions.

6. The medium of claim 5 , wherein:
one or more steps in the sequence are effectuated by executing a script encoding computer program operations for a browser instance to apply different types of discounts including at least one affiliate or tracking link and one discount code identified to the set of discounts by:
triggering the at least one affiliate or tracking link,
adding the one or more items of the given one of the constituent transactions to a cart with the merchant,
applying the discount code,
applying user shipping information based on the user record,
applying the one or more stored value cards, and
submitting the transaction with the merchant, and

the discount transaction is effectuated by iteratively processing the other constituent transactions in the set of constituent transactions.

7. The medium of claim 4 , wherein the at least one discount transaction is encoded in a data structure, the data structure comprising:
one or more values and functions for modifying the price and the estimated discount price of a set of constituent transactions of the at least one discount transaction based on a determined selection of items that may be added to the discount transaction or swapped with one or more of the items of the discount transaction, and
wherein an item swapped with one or more of the items of the discount transaction or added to the discount transaction is added to a one of the constituent transactions in the set of constituent transactions based on a determination that adding the item to the one of the constituent transactions provides a greater discount than adding the item to another constituent transaction in the set of constituent transactions.

8. The medium of claim 7 , wherein:
the determination that adding the item to the one of the constituent transactions provides a greater discount than adding the item to another constituent transaction in the set of constituent transactions is based on an output of at least one function including at least one of the values encoded in the data structure, the at least one of the values associated with the item or a merchant providing the item.

9. The medium of claim 1 , wherein determining an optimal set of discounts with a given merchant out of a plurality of other merchants, comprises:
determining whether any discounts having unmet criteria are available with the given merchant, the optimal set of discounts including at least one discount having unmet criteria by a discount transaction; and
determining a value corresponding to the unmet criteria is within a threshold amount and satisfying the unmet criteria, wherein satisfying the unmet criteria includes increasing a pre-discount cost of the transaction with the given merchant up to the threshold amount by selection of an addition or optioning of an item or a higher cost shipping option based on one or more of the user record, the item records, and the associated discounts or other item records having associated discounts.

10. The medium of claim 1 , wherein training, based at least in part on the

Assignees

• Retailmenot Inc

Inventors

• Kendall Joe
• Dubuque Shaun
• Sagi Vivek

Classifications

• G06Q20/3433

  the counter having monetary units · CPC title

• G06Q30/0239

  Online discounts or incentives · CPC title

• G06Q30/0222 Primary

  During e-commerce, i.e. online transactions · CPC title