Methods and apparatus for improving the selection of advertising
AssigneeWalmart Apollo, LLC
InventorsPeng Yang, Dong Xu, Yue Guo
Life of this patent
Estimating the term…
Abstract
The disclosed subject matter relates to a system and method for selecting/recommending ads based on a contextual bandit approach. The disclosed approach leverages various embedding vectors of item, search, page taxonomy trained based on traffic data via advanced deep learning models, and uses model signals (e.g. historical CTR, item price, rating, quality) from other ad placements. The learning mechanism on top of the current methodology to automatic chooses the best feature sets and adjust model performance over time. The contextual bandit model performs better with respect to CTR than the Thompson Sampling model, and achieves lower regret and faster convergence over time.
Claims20
A system for retrieving marketing ads implementing a contextual bandit in response to a user query comprising:
- a communication system;
- a database containing a plurality of marketing ads;
- a consumer device;
- and
- a computing device operably connected to the database, the consumer device, and the communication system, the computing device, using a deep learning model, configured to:
- assign a first parameter set for each of the plurality of marketing ads;
- receive an input query from a user of the consumer device;
- determine at least one feature of the query;
- filter the plurality of marketing ads from the database based on the at least one feature of the query to create a first set of candidate marketing ads;
- determine a context feature set parameter for each of the candidate marketing ads in the first set of candidate marketing ads, wherein the context feature set parameter is based on at least one feature of the first set of candidate marketing ads and is a function of at least the first parameter set;
- select a second set of marketing ads based upon the context feature set parameter of each of the candidate marketing ads in the first set;
- transmit the second set of marketing ads to the consumer device of the user;
- receive a reward metric based on user interaction with the second set of marketing ads;
- and
- update and optimize the first parameter set for each of the marketing ads in the second set as a function of the reward metric.
A method for improving a selection of marketing ads implementing a contextual bandit in response to a user query and using a deep learning model, comprising:
- assigning a first parameter set for each of a plurality of marketing ads;
- receiving an input query from a user of a consumer device;
- determining at least one feature of the query;
- filtering the plurality of marketing ads from a database based on the at least one feature of the query to create a first set of candidate marketing ads;
- determining a context feature set parameter for each of the candidate marketing ads in the first set of candidate marketing ads, wherein the context feature set parameter is based on at least one feature of the first set of candidate marketing ads and is a function of at least the first parameter set;
- selecting a second set of marketing ads based upon the context feature set parameter of each of the candidate marketing ads in the first set;
- transmitting the second set of marketing ads to the consumer device of the user;
- receiving a reward metric based on user interaction with the second set of marketing ads;
- and
- updating and optimizing the first parameter set for each of the marketing ads in the second set as a function of the reward metric.
A non-transitory computer readable medium having instructions stored thereon, wherein the instructions, when executed by at least one processor of a device for retrieving marketing ads implementing a contextual bandit in response to a user query and using a deep learning model, cause the device to perform operations comprising:
- assigning a first parameter set for each of a plurality of marketing ads;
- receiving an input query from a user of a consumer device;
- determining at least one feature of the query;
- filtering the plurality of marketing ads from a database based on the at least one feature of the query to create a first set of candidate marketing ads;
- determining a context feature set parameter for each of the candidate marketing ads in the first set of candidate marketing ads, wherein the context feature set parameter is based on at least one feature of the first set of candidate marketing ads and is a function of at least the first parameter set;
- selecting a second set of marketing ads based upon the context feature set parameter of each of the candidate marketing ads in the first set;
- transmitting the second set of marketing ads to the consumer device of the user;
- receiving a reward metric based on user interaction with the second set of marketing ads;
- and
- updating and optimizing the first parameter set for each of the marketing ads in the second set as a function of the reward metric.
Dependencies are inferred from the claim text.