Content-related actions based on context

1 . A system comprising:

one or more processors; and

one or more memories storing instructions that, upon execution by the one or more processors, configure the system to:

send, to a device, an object that causes a first presentation of first content and a second presentation of second content in association with the first presentation, the object indicating that the second content is to be requested;

receive, from the device based at least in part on the object and while the first presentation and the second presentation are caused, first data indicating a first request for the second content;

generate, based at least in part on the first data, contextual data indicating the first presentation by the device and including a content identifier of the first content;

generate an association between the contextual data and at least one of a device identifier of the device or an account identifier of an account associated with the device;

receive, from the device, second data corresponding to a natural language user input;

determine, by at least using a language model, that the second data indicates a second request for a content action, the content action determined based at least in part on the second data, the association, and the contextual data; and

cause the content action to be performed.

2 . The system of claim 1 , wherein the language model is configured to determine, upon the second data lacking an indication of target content on which the content action is to be performed, the content action and the target content based at least in part on the second data and the contextual data, wherein the content action is performed on the target content.

3 . The system of claim 1 , wherein the language model is configured to determine the content action based at least in part on the second data and target content on which the content action is to be performed based at least in part on the contextual data.

4 . The system of claim 1 , wherein the language model is configured to determine, based at least in part on the second data and the contextual data, a type of the content action and target content on which the content action is to be performed.

5 . The system of claim 1 , wherein the object is configured to cause the device to present the second content upon the first presentation of the first content and to request the second content upon the second presentation of the second content.

6 . The system of claim 1 , wherein the object indicates a parameter to present the second content in association with the first presentation of the first content and a frequency at which the second content is to be requested up on the second presentation of the second content.

7 . The system of claim 1 , wherein the object is configured to cause the device to increment a time counter upon the second presentation of the second content and to send the first data based at least in part on a value of the time counter.

8 . The system of claim 1 , wherein the second data includes a network address of the second content, and wherein the instructions, upon execution by the one or more processors, further configure the system to:

determine an association between the first request and the first content based at least in part on the network address, wherein the contextual data is generated further based at least in part on the association.

9 . The system of claim 1 , wherein the second data includes a network address of the second content and a value of a time counter associated with requesting the second content, wherein the value is appended to the network address.

10 . The system of claim 9 , wherein the instructions, upon execution by the one or more processors, further configure the system to:

determine, from the network address, the value; and

generate third data that includes a timestamp representing the value and a content identifier of the first content.

11 . The system of claim 1 , wherein the instructions, upon execution by the one or more processors, further configure the system to:

store, based at least in part on the first data, third data indicating the first presentation of the first content by the device;

determine that the second content has not been requested for a period of time subsequent to receiving the second data; and

indicate that the third data is invalid.

12 . The system of claim 11 , wherein the content action is a first content action, wherein the instructions, upon execution by the one or more processors, further configure the system to:

receive, while the third data is valid, fourth data indicating a third request for a second content action;

determine, based at least in part on the third data being valid, that the second content action is associated with the first presentation of the first content; and

cause the content action to be performed.

13 . A method comprising:

sending, to a device, an object that causes a first presentation of first content and a second presentation of second content in association with the first presentation, the object indicating that the second content is to be requested;

receiving, from the device based at least in part on the object and while the first presentation and the second presentation are caused, first data indicating a first request for the second content;

generating, based at least in part on the first data, contextual data indicating the first presentation by the device and including a content identifier of the first content;

generating an association between the contextual data and at least one of a device identifier of the device or an account identifier of an account associated with the device;

receiving, from the device, second data corresponding to a natural language user input;

determining, by at least using a language model, that the second data indicates a second request for a content action, the content action determined based at least in part on the second data, the association, and the contextual data; and

causing the content action to be performed.

14 . The method of claim 13 , further comprising:

determining an association between the first request and the first content based at least in part on source information of a content source, wherein the contextual data is generated based at least in part on the source information, and wherein the object includes the source information.

15 . The method of claim 13 , further comprising:

receiving, from the device based at least in part on the object and while the first presentation and the second presentation are caused, third data indicating a third request for the second content, wherein a time difference associated with the first data being received and the third data being received is indicated by the object.

16 . The method of claim 13 , further comprising:

receiving third data from another device, wherein the device and the other device are associated with an account, the third data indicating a third request for another content action;

determining that the third request is associated with the account;

determining, based at least in part on the contextual data, that the other content action is associated with the first presentation of the first content; and

causing the other content action to be performed by the other device.

17 . One or more non-transitory computer-readable storage media storing instructions that, upon execution on a system, cause the system to perform operations comprising:

sending, to a device, an object that causes a first presentation of first content and a second presentation of second content in association with the first presentation, the object indicating that the second content is to be requested;

receiving, from the device based at least in part on the object and while the first presentation and the second presentation are caused, first data indicating a first request for the second content;

generating, based at least in part on the first data, contextual data indicating the first presentation by the device and including a content identifier of the first content;

generating an association between the contextual data and at least one of a device identifier of the device or an account identifier of an account associated with the device;

receiving, from the device, second data corresponding to a natural language user input;

determining, by at least using a language model, that the second data indicates a second request for a content action, the content action determined based at least in part on the second data, the association, and the contextual data; and

causing the content action to be performed.

18 . The one or more non-transitory computer-readable storage media of claim 17 , wherein the operations further comprise:

associating a presentation identifier with a content identifier of the first content; and

including the presentation identifier in the object, wherein the first data indicates the presentation identifier, and wherein the contextual data is generated based at least in part on the presentation identifier.

19 . The one or more non-transitory computer-readable storage media of claim 18 , wherein the operations further comprise:

determining the presentation identifier from the first data;

determining the content identifier based at least in part on an association between the presentation identifier and the content identifier; and

storing the content identifier in the contextual data.

20 . The one or more non-transitory computer-readable storage media of claim 17 , wherein the operations further comprise:

associating a presentation identifier with the content identifier of the first content and at least one of the device identifier of the device or the account identifier of an account associated with the device;

determining the presentation identifier from the first data; and

determining the content identifier based at least in part on an association between the presentation identifier and the content identifier.