PARAMETERIZING NETWORK COMMUNICATION PATHS

1 . A system to parameterize network communication paths, comprising:

a data processing system including a network path detection component, a parameterization component, and a content selector component executed by at least one processor, the data processing system to:

identify a network communication path traversed by a plurality of client computing devices via a network protocol, the network communication path including a first web page and a second web page, the first web page having a content item with a hyperlink that directs the plurality of client computing devices to the second web page;

generate, based on a parameter of the network communication path and an attribute of each of a plurality of keywords to the hyperlink on the first web page that directed the plurality of client computing devices to the second web page via the network protocol, a weight data structure with a scalar value for each of the plurality of keywords; and

select, based on a comparison of the scalar value for each of the plurality of keywords, a keyword of the plurality of keywords to input into an online content selection process to select the content item in response to a request for content received from a client computing device.

2 . The system of claim 1 , wherein the first web page comprising an anchor web page, and the second web page comprising a landing web page provided by a content provider device.

3 . The system of claim 1 , comprising the data processing system to scrape the first web page to identify the plurality of keywords, and store the plurality of keywords in a data repository.

4 . The system of claim 1 , comprising the data processing system to:

render code of the first web page;

determine a distance between each of the plurality of keywords of the first web page and the content item comprising the hyperlink; and

determine the scalar value for each of the plurality of keywords based on the distance between each of the plurality of keywords of the first web page and the content item.

5 . The system of claim 4 , wherein the attribute includes a proximity, comprising the data processing system to:

determine a distance between a first keyword of the plurality of keywords of the first web page and the content item based on a number of characters between the first keyword and the content item; and

determine the scalar value for each of the first keyword based on the distance between the first keyword and the content item.

6 . The system of claim 4 , comprising the data processing system to:

determine a distance between a first keyword of the plurality of keywords of the first web page and the content item based on a number of pixels between the first keyword and the content item; and

determine the scalar value for each of the first keyword based on the distance between the first keyword and the content item.

7 . The system of claim 4 , comprising the data processing system to:

determine a bucketized distance between a first keyword of the plurality of keywords of the first web page and the content item, the bucketized distance corresponding to at least one of:

within a threshold number of words;

within a same sentence;

within a same paragraph; and

within a same fold of the first web page; and

determine the scalar value for each of the first keyword based on the distance between the first keyword and the content item.

8 . The system of claim 1 , comprising the data processing system to:

identify a second network communication path traversed by a second plurality of client computing devices via the network protocol, the second network communication path including a third web page and the second web page, the third web page having the content item with the hyperlink that directs the second plurality of client computing devices to the second web page;

generate, based on a second parameter of the second network communication path and a proximity of each of a second plurality of keywords to the hyperlink on the third web page that directed the second plurality of client computing devices to the second web page via the network protocol, the weight data structure with the scalar value for each of the second plurality of keywords.

9 . The system of claim 8 , comprising the data processing system to:

determine that the second parameter of the second network communication path is greater than the parameter of the network communication path; and

assign a greater weight to the second plurality of keywords based on the second parameter greater than the first parameter.

10 . The system of claim 1 , comprising the data processing system to:

generate the weight data structure with the scalar value for each of the plurality of keywords based on:

the parameter of the network communication path,

a proximity of each of the plurality of keywords to the hyperlink on the first web page that directed the plurality of client computing devices to the second web page via the network protocol, and

a network activity on the second web page.

11 . The system of claim 1 , wherein the parameter comprises a total number of instances of the first web page directing the plurality of client computing devices to the second web page during a time interval.

12 . The system of claim 1 , wherein the parameter comprises a total number of instances of the first web page directing the plurality of client computing devices to the second web page divided by a total number of instances of a plurality of web pages directing a second plurality of client computing devices to the second web page during a time interval.

13 . The system of claim 1 , comprising the data processing system to:

determine a second attribute associated with a cluster formed of at least two of the plurality of client computing devices; and

select the keyword of the plurality of keywords to input into the online content selection process to select the content item in response to the request for content received from the client computing device matching the second attribute associated with the cluster.

14 . A method of parameterizing network communication paths, comprising:

identifying, by a network path detection component executed by a data processing system comprising at least one processor, a network communication path traversed by a plurality of client computing devices via a network protocol, the network communication path including a first web page and a second web page, the first web page having a content item with a hyperlink that directs the plurality of client computing devices to the second web page;

generating, by a parameterization component executed by the data processing system, based on a parameter of the network communication path and an attribute of each of a plurality of keywords to the hyperlink on the first web page that directed the plurality of client computing devices to the second web page via the network protocol, a weight data structure with a scalar value for each of the plurality of keywords; and

selecting, by a content selector component executed by the data processing system, based on a comparison of the scalar value for each of the plurality of keywords, a keyword of the plurality of keywords to input into an online content selection process to select the content item in response to a request for content received from a client computing device.

15 . The method of claim 14 , wherein the first web page comprising an anchor web page, and the second web page comprising a landing web page provided by a content provider device.

16 . The method of claim 14 , comprising:

scraping, by the data processing system, the first web page to identify the plurality of keywords, and store the plurality of keywords in a data repository.

17 . The method of claim 14 , wherein the attribute comprises a proximity, the method comprising:

rendering, by the data processing system, code of the first web page;

determining, by the data processing system, a distance between each of the plurality of keywords of the first web page and the content item comprising the hyperlink; and

determining, by the data processing system, the scalar value for each of the plurality of keywords based on the distance between each of the plurality of keywords of the first web page and the content item.

18 . The method of claim 17 , wherein the attribute comprises a proximity, the method comprising:

determining, by the data processing system, a distance between a first keyword of the plurality of keywords of the first web page and the content item based on a number of characters between the first keyword and the content item; and

determining, by the data processing system, the scalar value for each of the first keyword based on the distance between the first keyword and the content item.

19 . The method of claim 17 , wherein the attribute comprises a proximity, the method comprising:

determining, by the data processing system, a distance between a first keyword of the plurality of keywords of the first web page and the content item based on a number of pixels between the first keyword and the content item; and

determining, by the data processing system, the scalar value for each of the first keyword based on the distance between the first keyword and the content item.

20 . The method of claim 17 , wherein the attribute comprises a proximity, the method comprising:

determining, by the data processing system, a bucketized distance between a first keyword of the plurality of keywords of the first web page and the content item, the bucketized distance corresponding to at least one of:

within a threshold number of words;

within a same sentence;

within a same paragraph; and

within a same fold of the first web page; and

determining, by the data processing system, the scalar value for each of the first keyword based on the distance between the first keyword and the content item.