Back to results
US 11,900,110 B2Granted
Increasing user interaction with deep learning agent
AssigneeProsper Funding LLC
InventorPaul Golding
Life of this patent
Estimating the term…
Abstract
A method increases user interaction with deep learning agents. A generation request for a content slot is received. A subsequent tag vector, for the content slot, is generated from a previous tag, for a previous content slot, using a subsequent tag model. A context vector is generated from a set of subsequent tag vectors, which include the subsequent tag vector, using a context vector generator. A selection vector is generated from the context vector using a contextual bandit model. A content set is generated for a content slot using the selection vector. The content set for the content slot is presented.
Claims18
A method comprising:
- receiving a generation request for a content slot;
- generating a subsequent tag vector, for the content slot, from a previous tag, for a previous content slot, using a subsequent tag model;
- generating a context vector from a set of subsequent tag vectors, which include the subsequent tag vector, using a context vector generator;
- generating a selection vector from the context vector using a contextual bandit model;
- generating a content set for the content slot using the selection vector and a content template, wherein generating the content set comprises:
- selecting the content template using a random number and a plurality of values from the selection vector, wherein the plurality of values is a plurality of probabilities used to select the content template from a plurality of content templates mapped to the plurality of probabilities;
- and
- presenting the content set for the content slot.
5 more elements
A system comprising:
- a server comprising one or more processors and one or more memories;
- and
- a server application, executing on the one or more processors of the server, configured for:
- receiving a generation request for a content slot,
- generating a subsequent tag vector, for the content slot, from a previous tag, for a previous content slot, using a subsequent tag model,
- generating a context vector from a set of subsequent tag vectors, which include the subsequent tag vector, using a context vector generator,
- generating a selection vector from the context vector using a contextual bandit model,
- generating a content set for the content slot using the selection vector and a content template, wherein generating the content set comprises:
- selecting the content template using a random number and a plurality of values from the selection vector, wherein the plurality of values is a plurality of probabilities used to select the content template from a plurality of content templates mapped to the plurality of probabilities, and
- presenting the content set for the content slot.
7 more elements
A method comprising:
- training a subsequent tag model to generate a subsequent tag vector from a previous tag;
- training a plurality of contextual bandit models, comprising a contextual bandit model, to each generate selection vectors from context vectors;
- generating the subsequent tag vector, fora content slot, from the previous tag, for a previous content slot, using the subsequent tag model;
- generating a context vector from a set of subsequent tag vectors, which include the subsequent tag vector, using a context vector generator;
- generating a selection vector from the context vector using the contextual bandit model;
- and
- generating a content set for the content slot using the selection vector and a content template, wherein generating the content set comprises:
- selecting the content template using a random number and a plurality of values from the selection vector, wherein the plurality of values is a plurality of probabilities used to select the content template from a plurality of content templates mapped to the plurality of probabilities.
5 more elements
Dependencies are inferred from the claim text.