AUTOMATED CONTENT GENERATION AND DELIVERY
Automated content generation and delivery may include processing a request for story synthesis using specified content items. The request may, for example, be captured using a microphone of an electronic device and transmitted to a server device. The specified content items may be mapped to a story template. Based on the story template, other content items related to the specified content items may be retrieved from one or more content sources. The content sources may, for example, refer to websites, social media platforms, search engine results, or other data stores. A story may then be synthesized using the specified content items and the other content items, for example, by combining the specified content items and the other content items according to the story template. The synthesized story may then be output, for example, at the electronic device.
1 - 20 . (canceled)
21 . A method, comprising:
receiving a request to create a story synthesis that simulates an answer to a query;
inputting the received request to create the story synthesis into a machine learning model to generate an output;
determining a stage of processing of the machine learning model, from a plurality of stages of processing, at a time when the request to create the story synthesis was inputted;
generating an output from the machine learning model that is based on the determined stage of processing of the machine learning model;
creating a structure for the story synthesis based on the generated output from the machine learning model;
obtaining story components to populate the created story structure; and
stitching together the story components based on the created structure to output the answer to the query.
22 . The method of claim 21 , wherein the stage of processing at the time when the request to create the story synthesis was inputted into the machine learning model is a stage of using specified content items to select a story template; and based on the stage of processing, generating the output from the machine learning model that is a selection of a story template.
23 . The method of claim 22 , further comprising:
receiving a selection of the story template; and
using the machine learning model to determine which types of other content items to use based on selection of the story template.
24 . The method of claim 23 , further comprising:
determining that one or more types of other content items to use based on the selection of the story template is not available; and
in response to determining that the one or more types of other content items to use based on the selection of the story template is not available, crawling one or more content sources to obtain the one or more types of other content items.
25 . The method of claim 23 , further comprising:
determining that one or more types of other content items to use based on the selection of the story template is not available; and
in response to determining that the one or more types of other content items to use based on the selection of the story template is not available:
analyzing previously used content items by the machine learning model to determine their contextual relevance to the story template; and
using the previously used content items to populate the created story structure upon determining that the previously used content items are contextually relevant to the story template.
26 . The method of claim 21 , further comprising:
generating a verification prompt, wherein the verification prompt includes information relating to some or all portions of the stitched story components and seeks user confirmation; and
presenting the verification prompt to a user from whom the request was received.
27 . The method of claim 26 , further comprising:
receiving user confirmation in response to the presented verification prompt; and
associating the user confirmation with some or all portions of the stitched story components presented in the verification prompt as being relevant to the received request.
28 . The method of claim 26 , further comprising:
determining based on a response to the verification prompt that includes information relating to some or all portions of the stitched story components does not match the received request; and
in response to determining that the information does not match the received request, transmitting a signal to an electronic device of the user to cause the electronic device to prompt for further input.
29 . The method of claim 21 , further comprising:
determining whether the received request includes a query that is specific to a geographic location; and
in response to determining that the received request includes a query that is specific to a geographic location, relating story synthesis to the geographic location.
30 . The method of claim 21 , further comprising, organizing the machine learning model into an input later, intermediary layer, and an output layer, wherein
the input layer is used for organizing the request to create the story synthesis,
the intermediary layer is used for performing convolution-and-pooling, and
the output layer is used for generating the output based on the determined stage of processing of the machine learning model.
31 . A system comprising:
a processor configured to:
receive a request to create a story synthesis that simulates an answer to a query;
input the received request to create the story synthesis into a machine learning model to generate an output;
determine a stage of processing of the machine learning model, from a plurality of stages of processing, at a time when the request to create the story synthesis was inputted;
generate an output from the machine learning model that is based on the determined stage of processing of the machine learning model;
create a structure for the story synthesis based on the generated output from the machine learning model;
obtain story components to populate the created story structure; and
stitch together the story components based on the created structure to output the answer to the query.
32 . The system of claim 31 , wherein the stage of processing at the time when the request to create the story synthesis was inputted into the machine learning model is a stage of using specified content items to select a story template; and based on the stage of processing, generating the output from the machine learning model that is a selection of a story template.
33 . The system of claim 32 , further comprising, the processor configured to:
receive selection of the story template; and
use the machine learning model to determine which types of other content items to use based on selection of the story template.
34 . The system of claim 33 , further comprising, the processor configured to:
determine that one or more types of other content items to use based on the selection of the story template is not available; and
in response to determining that the one or more types of other content items to use based on the selection of the story template is not available, crawl one or more content sources to obtain the one or more types of other content items.
35 . The system of claim 33 , further comprising, the processor configured to:
determine that one or more types of other content items to use based on the selection of the story template is not available; and
in response to determining that the one or more types of other content items to use based on the selection of the story template is not available:
analyze previously used content items by the machine learning model to determine their contextual relevance to the story template; and
use the previously used content items to populate the created story structure upon determining that the previously used content items are contextually relevant to the story template.
36 . The system of claim 31 , further comprising, the processor configured to:
generate a verification prompt, wherein the verification prompt includes information relating to some or all portions of the stitched story components and seeks user confirmation; and
present the verification prompt to a user from whom the request was received.
37 . The system of claim 36 , further comprising, the processor configured to:
receive user confirmation in response to the presented verification prompt; and
associate the user confirmation with some or all portions of the stitched story components presented in the verification prompt as being relevant to the received request.
38 . The system of claim 36 , further comprising, the processor configured to:
determine based on a response to the verification prompt that includes information relating to some or all portions of the stitched story components does not match the received request; and
in response to determining that the information does not match the received request, transmit a signal to an electronic device of the user to cause the electronic device to prompt for further input.
39 . The system of claim 31 , further comprising, the processor configured to:
determine whether the received request includes a query that is specific to a geographic location; and
in response to determining that the received request includes a query that is specific to a geographic location, relate story synthesis to the geographic location.
40 . The system of claim 31 , further comprising, the processor configured to organize the machine learning model into an input later, intermediary layer, and an output layer, wherein
the input layer is used for organizing the request to create the story synthesis,
the intermediary layer is used for performing convolution-and-pooling, and
the output layer is used for generating the output based on the determined stage of processing of the machine learning model.