Predictive recommendation system using price boosting
View Patent ↗In general, embodiments of the present invention provide systems, methods and computer readable media for ranking promotions selected for recommendation to consumers based on predictions of promotion performance and consumer behavior. In embodiments, a set of promotions to be recommended to a consumer can be sorted and/or ranked according to respective relevance scores representing a probability that the consumer's behavior in response to the promotion will match a ranking target. In embodiments, calculating scores is based on a relevance model (a predictive function) derived from one or more contextual data sources representing attributes of promotions and consumer behavior. In embodiments, an absolute relevance score represents an absolute prediction of a ranking target variable. In embodiments, absolute relevance may be used to determine personalized local merchant discovery frontiers; featured result set thresholding for impressions; and/or promotion notification triggers. In embodiments, predictive models based on gross revenue may be optimized using promotion category-dependent price boosting.
1. A computer-implemented method, comprising:
receiving, using a processor, a set of available promotions for a consumer associated with a consumer device, each available promotion being associated with a relevance score representing an estimated conversion rate per impression for the promotion, wherein the relevance score is based on a relevance model, wherein the relevance model is a predictive function;
ranking, using the processor, the available promotions based on their respective relevance scores;
adjusting, by the processor, each of the relevance scores to represent an estimated minimum quantity of bookings per impression for the promotion associated with the score using price boosting;
re-ranking, by the processor, the available promotions based on their respective adjusted scores; and
presenting a subset comprising a top-N ranked promotions of the re-ranked available promotions to the consumer device;
wherein adjusting, by the processor, each of the relevance scores comprises:
determining a level of engagement for the consumer; and
adjusting each of the relevance scores in an instance in which the consumer is at a highly engaged level.
2. The method of claim 1 , wherein adjusting the relevance score comprises:
receiving features representing the promotion associated with the score, wherein the features include the promotion price and the promotion category;
comparing the promotion price to a price threshold associated with the promotion category; and
in an instance in which the promotion price is greater than the price threshold,
capping the promotion price by assigning the price threshold as the promotion price.
3. The method of claim 1 , further comprising:
adjusting the relevance score by multiplying the relevance score by a conversion rate multiplier based on the promotion price.
4. The method of claim 1 , wherein the price threshold associated with the promotion category is determined based on configurable parameters including the price point below which the price adjustment is applied to the conversion rate, the lowest price point for the promotion category, and a price exponent.
5. The method of claim 1 , wherein the price threshold associated with the promotion category is determined based on at least one of business strategy, promotion data describing attributes of promotion performance, and user data describing attributes of consumer behavior.
6. The method of claim 5 , wherein promotions belonging to the promotion category are to be de-emphasized, and wherein determining the price threshold based on business strategy comprises:
identifying an ideal price threshold value by selecting a price at which the bulk of the bookings are generated for the promotion category; and
setting the promotion category price threshold to a value that is less than the ideal price threshold value.
7. The method of claim 1 , further comprising:
not adjusting, using the processor, each of the relevance scores in an instance in which the consumer is not at a highly engaged level.
8. A computer program product, encoded on a computer-readable medium, operable to cause data processing apparatus to perform operations comprising:
receiving a set of available promotions for a consumer associated with a consumer device, each available promotion being associated with a relevance score representing an estimated conversion rate per impression for the promotion, wherein the relevance score is based on a relevance model, wherein the relevance model is a predictive function;
ranking the available promotions based on their respective relevance scores;
adjusting each of the relevance scores to represent an estimated minimum quantity of bookings per impression for the promotion associated with the score using price boosting;
re-ranking the available promotions based on their respective adjusted scores; and
presenting a subset comprising a top-N ranked promotions of the re-ranked available promotions to the consumer device;
wherein adjusting each of the relevance scores comprises:
determining a level of engagement for the consumer; and
adjusting each of the relevance scores in an instance in which the consumer is at a highly engaged level.
9. The computer program product of claim 8 ,
wherein adjusting, using the processor, the relevance score comprises:
receiving features representing the promotion associated with the score, wherein the features include the promotion price and the promotion category;
comparing the promotion price to a price threshold associated with the promotion category; and
in an instance in which the promotion price is greater than the price threshold,
capping the promotion price by assigning the price threshold as the promotion price.
10. The computer program product of claim 9 , further comprising:
adjusting, using the processor, the relevance score by multiplying the relevance score by a conversion rate multiplier based on the promotion price.
11. The computer program product of claim 9 , wherein the price threshold associated with the promotion category is determined based on configurable parameters including the price point below which the price adjustment is applied to the conversion rate, the lowest price point for the promotion category, and a price exponent.
12. The computer program product of claim 9 , wherein the price threshold associated with the promotion category is determined based on at least one of business strategy, promotion data describing attributes of promotion performance, and user data describing attributes of consumer behavior.
13. The computer program product of claim 12 , wherein
promotions belonging to the promotion category are to be de-emphasized, and wherein determining the price threshold based on business strategy comprises:
identifying, using the processor, an ideal price threshold value by selecting a price at which the bulk of the bookings are generated for the promotion category; and
setting, using the processor, the promotion category price threshold to a value that is less than the ideal price threshold value.
14. The computer program product of claim 8 , further comprising:
not adjusting each of the relevance scores in an instance in which the consumer is not at a highly engaged level.
15. A system, comprising:
one or more computers and one or more storage devices storing instructions that are operable, when executed by the one or more computers, to cause the one or more computers to perform operations comprising:
receiving a set of available promotions for a consumer associated with a consumer device, each available promotion being associated with a relevance score representing an estimated conversion rate per impression for the promotion, wherein the relevance score is based on a relevance model, wherein the relevance model is a predictive function;
ranking the available promotions based on their respective relevance scores;
adjusting each of the relevance scores to represent an estimated minimum quantity of bookings per impression for the promotion associated with the score using price boosting;
re-ranking the available promotions based on their respective adjusted scores; and
presenting a subset comprising a top-N ranked promotions of the re-ranked available promotions to the consumer device;
wherein adjusting each of the relevance scores comprises:
determining a level of engagement for the consumer; and
adjusting each of the relevance scores in an instance in which the consumer is at a highly engaged level.
16. The system of claim 15 , wherein adjusting the relevance score comprises:
receiving features representing the promotion associated with the score, wherein the features include the promotion price and the promotion category;
comparing the promotion price to a price threshold associated with the promotion category; and
in an instance in which the promotion price is greater than the price threshold,
capping the promotion price by assigning the price threshold as the promotion price.
17. The system of claim 16 , further comprising:
adjusting the relevance score by multiplying the relevance score by a conversion rate multiplier based on the promotion price.
18. The system of claim 16 , wherein the price threshold associated with the promotion category is determined based on configurable parameters including the price point below which the price adjustment is applied to the conversion rate, the lowest price point for the promotion category, and a price exponent.
19. The system of claim 16 , wherein the price threshold associated with the promotion category is determined based on at least one of business strategy, promotion data describing attributes of promotion performance, and user data describing attributes of consumer behavior.
20. The system of claim 19 , wherein promotions belonging to the promotion category are to be de-emphasized, and wherein determining the price threshold based on business strategy comprises:
identifying an ideal price threshold value by selecting a price at which the bulk of the bookings are generated for the promotion category; and
setting the promotion category price threshold to a value that is less than the ideal price threshold value.
21. The system of claim 15 , further comprising:
not adjusting each of the relevance scores in an instance in which the consumer is not at a highly engaged level.
22. The system of claim 15 , wherein the predictive function is generated offline using supervised machine learning in a set of modeling stages in which the predictive function is adapted based on training data sets of features that are extracted from a set of data sources.
23. The system of claim 15 , wherein the predictive function is an ensemble of trees and a definition of the predictive function is represented in Predictive Model Markup Language (PMML).
24. The system of claim 23 , wherein the predictive function definition is a file comprising one or more Predictive Model Markup Language (PMML) model definitions and a decision tree for use in selecting a predictive model definition from the one or more Predictive Model Markup Language (PMML) model definitions.
25. The system of claim 23 , wherein the file is a JavaScript Object Notation (JASON) file.
26. The system of claim 23 , wherein the predictive function includes multiple random forest models, each representing a different contextual data source.