US 11,960,517 B2Granted

Dynamic cross-platform ask interface and natural language processing model

AssigneeAdobe Inc.

InventorsPiyush Gupta, Binit Kumar Sinha, Eunyee Koh, Fan Du, Gaurav Makkar, Silky Kedawat, Subrahmanya Kumar Giliyaru, Vasanthi Holtcamp, Nikhil Belsare

Life of this patent

Estimating the term…

patentgrep

Loading patent…

Abstract

The present disclosure relates to systems, non-transitory computer-readable media, and methods that generate a dynamic cross-platform ask interface and utilize a cross-platform language processing model to provide platform-specific, contextually based responses to natural language digital text queries. In particular, in one or more embodiments, the disclosed systems utilize machine learning models to extract registered intents from digital text queries to identify platform-specific configurations associated with the registered intents. Utilizing the platform-specific configurations, the disclosed systems can generate tailored platform-specific requests for information, as well as customized end-user search results that cause client devices to efficiently, accurately, and flexibly render platform-specific search results.

Claims20

Dependent claims

A non-transitory computer-readable storage medium comprising instructions that, when executed by at least one processor, cause a computing device to:

register a plurality of software platforms to identify intents corresponding to the plurality of software platforms by training a cross-platform language processing model with:
a first set of previous digital text queries for a first software platform, a first set of ground truth intents for the first software platform, and a first set of platform-specific configurations for the first software platform;
and
a second set of previous digital text queries for a second software platform, a second set of ground truth intents for the second software platform, and a second set of platform-specific configurations for the second software platform;
receive, via a query interface of a client device, a digital text query for information from a software platform;
determine, utilizing the cross-platform language processing model trained for the plurality of software platforms, at least one intent from the digital text query;
select, utilizing the at least one intent, a plurality of platform-specific configurations corresponding to a plurality of software platforms;
generate a plurality of platform-specific requests utilizing the at least one intent and the plurality of platform-specific configurations;
transmit the plurality of platform-specific requests to the plurality of software platforms;
receive a plurality of responses to the plurality of platform-specific requests from the plurality of software platforms;
generate a response to the digital text query from the plurality of responses to the plurality of platform-specific requests;
and
cause the client device to display, via the query interface, the response to the digital text query.

10 more elements

A system comprising:

at least one computer memory device comprising:
a cross-platform language processing model, and
a plurality of platform-specific configurations corresponding to a plurality of software platforms and referencing a plurality of registered intents;
and
one or more servers configured to cause the system to:
register the plurality of software platforms to identify intents corresponding to the plurality of software platforms by training the cross-platform language processing model with:
a first set of previous digital text queries for a first software platform, a first set of ground truth intents for the first software platform, and a first set of platform-specific configurations for the first software platform;
and
a second set of previous digital text queries for a second software platform, a second set of ground truth intents for the second software platform, and a second set of platform-specific configurations for the second software platform;
receive, via a query interface of a software platform of a client device, a digital text query for information;
utilize the cross-platform language processing model trained for the plurality of software platforms to determine at least one intent from the digital text query;
select, utilizing the at least one intent, a subset of platform-specific configurations for a subset of software platforms from the plurality of platform-specific configurations corresponding to the plurality of software platforms;
identify a plurality of response rendition types from the subset of platform-specific configurations based on the at least one intent;
generate, utilizing the at least one intent and the plurality of platform-specific configurations, a plurality of platform-specific requests for transmission to the subset of software platforms;
and
generate, from a plurality of responses to the plurality of platform-specific requests, a response to the digital text query by combining the plurality of responses to the plurality of platform-specific requests and the plurality of response rendition types.

13 more elements

A computer-implemented method comprising:

generating, utilizing a cross-platform language processing model trained for a plurality of software platforms, at least one intent from a digital text query for information from a software platform, wherein the digital text query is received via a query interface of a client device;
generating a plurality of platform-specific requests from a plurality of platform-specific configurations corresponding to the at least one intent by:
registering the plurality of software platforms to identify intents corresponding to the plurality of software platforms by training the cross-platform language processing model with:
a first set of previous digital text queries for a first software platform, a first set of ground truth intents for the first software platform, and a first set of platform-specific configurations for the first software platform;
and
a second set of previous digital text queries for a second software platform, a second set of ground truth intents for the second software platform, and a second set of platform-specific configurations for the second software platform;
generating, from the plurality of platform-specific requests, a plurality of responses to the digital text query;
and
transmitting the plurality of responses to the digital text query to the client device such that the client device renders the plurality of responses to the digital text query by expanding the query interface into a response rendition canvas and rendering the plurality of responses to the digital text query within the response rendition canvas according to at least one of a table response rendition type, a reporting chart response rendition type, or a list response rendition type.

6 more elements

Dependencies are inferred from the claim text.