Apparatus and method for enhanced message targeting

Patent metadata

Field	Value
Publication number	US-11676162-B2
Application number	US-202117215968-A
Country	US
Kind code	B2
Filing date	Mar 29, 2021
Priority date	Mar 30, 2015
Publication date	Jun 13, 2023
Grant date	Jun 13, 2023

Abstract

A method, apparatus, and computer program product are disclosed for improved machine learning using a statistical model. In the context of an apparatus, some example embodiments include a processor configured to cause retrieval of information regarding a plurality of consumers, and modeling circuitry configured to train a statistical model of the plurality of consumers based on the retrieved information, and predict, using the statistical model, an incremental booking value associated with the promotion for each consumer of the plurality of consumers. The processor is further configured to select a subset of the plurality of consumers for receiving impressions of the promotion. Some example embodiments may further include communications circuitry configured to transmit an impression of the promotion to each consumer in the subset of the plurality of consumers.

First claim

What is claimed is:

1. An apparatus comprising at least one processor and at least one memory having computer-coded instructions stored thereon that, in execution with the at least one processor, configures the apparatus to:
train a statistical model using historical data retrieved from at least one database of a service, wherein the historical data comprises information regarding at least one transaction associated with each identifier entity of a plurality of identifier entities, and wherein each identifier entity is associated with one or more consumer devices used by a user to access the service, and wherein the historical data comprises a number of interactions associated with the identifier entity within at least one time period;
estimate a first incremental booking value associated with a first promotion for the consumer using the statistical model, wherein to estimate the first incremental booking value the apparatus is at least configured to use the statistical model to:
estimate a first expected booking value associated with at least a consumer for a first promotion, the first expected booking value based at least in part on first input information comprising a first indicator representing the consumer having access to the first promotion;
estimate a second expected booking value associated with at least the consumer for the first promotion, the second expected booking value based at least in part on second input information comprising a second indicator representing the consumer not having access to the first promotion;
calculate the first incremental booking value associated with the first promotion for the consumer based at least in part on the first expected booking value and the second expected booking value; and

select, using the statistical model, the consumer to receive a generated impression of the first promotion based at least in part on the first incremental booking value associated with the first promotion for the consumer.

2. The apparatus according to claim 1 , wherein the apparatus is configured to identify whether to select the consumer to receive the first promotion further based at least in part on a target parameter.

3. The apparatus according to claim 1 , wherein to train the statistical model using the historical data the apparatus is caused to:
train the statistical model based at least in part on a first subset of the historical data associated with the first indicator representing at least a portion of the plurality of identifier entities having access to at least one promotion and a second set of input information associated with the second indicator representing at least the portion of the plurality of identifier entities not having access to the at least one promotion.

4. The apparatus according to claim 1 , the apparatus further configured to:
generate, using the statistical model, an expected booking value set associated with the consumer for a promotion set, the expected booking value set comprising an expected booking value for each promotion in the promotion set, each additional expected booking value associated with the first indicator representing the consumer having access to the promotion,
wherein the promotion set comprises the first promotion, and
wherein the expected booking value set comprises the first expected booking value associated with the first promotion; and
determine an incremental booking value set associated with the promotion set, the incremental booking value set comprising an incremental booking value for each promotion in the promotion set for the consumer based at least in part on the expected booking value set and the second expected booking value,
wherein the apparatus is configured to identify whether to select the consumer to receive the first promotion based at least in part on the incremental booking value set associated with the promotion set.

5. The apparatus according to claim 4 , wherein to identify whether to select the consumer to receive the first promotion, the apparatus is configured to:
determine, from the promotion set, a promotion subset based on the incremental booking value set, the promotion subset comprising at least one promotion associated with an incremental booking value satisfying an incremental booking value threshold.

6. The apparatus according to claim 4 , wherein the promotion set comprises at least one promotion determined relevant to the consumer based on electronic marketing information associated with the consumer.

7. A computer-implemented method comprising:
training a statistical model using historical data retrieved from at least one database of a service, wherein the historical data comprises information regarding at least one transaction associated with each identifier entity of a plurality of identifier entities, and wherein each identifier entity is associated with one or more consumer devices used by a user to access the service, and wherein the historical data comprises a number of interactions associated with the identifier entity within at least one time period;
estimating a first incremental booking value associated with a first promotion for the consumer using the statistical model by at least:
estimating a first expected booking value associated with at least a consumer for a first promotion, the first expected booking value based at least in part on first input information comprising a first indicator representing the consumer having access to the first promotion;
estimating a second expected booking value associated with at least the consumer for the first promotion, the second expected booking value based at least in part on second input information comprising a second indicator representing the consumer not having access to the first promotion;
calculating the a first incremental booking value associated with the first promotion for the consumer based at least in part on the first expected booking value and the second expected booking value; and

selecting, using the statistical model, the consumer to receive a generated impression of the first promotion based at least in part on the first incremental booking value associated with the first promotion for the consumer.

8. The computer-implemented method according to claim 7 , wherein identifying whether to select the consumer to receive the first promotion is further based at least in part on a target parameter.

9. The computer-implemented method according to claim 7 , wherein training the statistical model using the historical data comprises:
training the statistical model based at least in part on a first subset of the historical data associated with the first indicator representing at least a portion of the plurality of identifier entities having access to at least one promotion and a second set of input information associated with the second indicator representing at least the portion of the plurality of identifier entities not having access to the at least one promotion.

10. The computer-implemented method according to claim 7 , the computer-implemented method further comprising:
generating, using the statistical model, an expected booking value set associated with the consumer for a promotion set, the expected booking value set comprising an expected booking value for each promotion in the promotion set, each additional expected booking value associated with the first indicator representing the consumer having access to the promotion,
wherein the promotion set comprises the first promotion, and
wherein the expected booking value set comprises the first expected booking value associated with the first promotion; and
determining an incremental booking value set associated with the promotion set, the incremental booking value s

Assignees

• Groupon Inc

Inventors

• Parekh Rajesh Girish
• Balakrishnan Raju
• Zhang Guoxian
• Gaurav Prashant

Classifications

• G06N20/00

  Machine learning · CPC title

• G06Q30/0211

  Determining the effectiveness of discounts or incentives · CPC title

• G06N20/20

  Ensemble learning · CPC title

• G06Q30/0201 Primary

  Market modelling; Market analysis; Collecting market data · CPC title