Systems and methods for advertising on content-screened web pages

A method for delivering advertising to publishers of content sites, the method comprising the following operations performed by one or more processors:

• receiving an impression request from a Uniform Resource Locator (“URL”) that is related to a content site, wherein the impression request includes a document referring address (“DREF”) parameter;
• determining whether the URL has one or more existing ratings stored in a response cache;
• appending, upon determining that the URL has one or more existing ratings stored in the response cache, the one or more existing ratings to the impression request;
• counting, upon determining that the URL does not have one or more existing ratings stored in the response cache, a number of impression requests received for the URL;
• determining whether the number of impression requests for the URL has reached a threshold number of requests;
• screening, by a content categorization service (“CCS”) engine, upon determining that the number of impression requests for the URL has reached the threshold number of requests, HTML content of the URL for one or more categories of content, the screening including fetching the HTML content from the URL;
• generating one or more ratings for the URL based on the screened HTML content, the one or more ratings comprising at least one rating for each of the one or more categories of content;
• storing the one or more ratings for the site URL in a response cache server;
• sending, to an ad server, the impression request with the one or more existing ratings, in a case in which the URL has the existing ratings stored in the response cache;
• serving a content-sensitive advertising impression for the content site based on (i) the existing one or more ratings, in a case in which the URL has the existing one or more ratings stored in the response cache, or (ii) the generated one or more ratings, in a case in which one or more ratings are generated for the URL;
• and
• serving a non-content-sensitive advertising impression for the content site, in a case in which the URL has no existing ratings stored in the response cache and the number of impression requests for the URL has not reached the threshold number of requests.

A method for content-screening to control delivering advertising to publishers of web pages, the method comprising the following operations performed by one or more processors:

• appending, upon determining that a Uniform Resource Locator (“URL”) related to a web site has one or more existing ratings stored in a response cache, the one or more existing ratings to a received impression request for the web page by extracting a document referring address (“DREF”) parameter embedded in the received impression request, and using the DREF parameter to look up the URL in the response cache;
• adding, upon determining that the URL has no existing ratings stored in the response cache, a request for the URL to a request queue;
• determining, in a case in which the request for the URL is added to the request queue, whether a number of requests for the URL in the request queue has reached a threshold number of requests;
• screening, by a content categorization service (“CCS”) engine, upon determining that the number of requests for the URL has reached the threshold number of requests, HTML content of the URL for one or more categories of content, the screening including fetching the HTML content from the URL;
• generating one or more ratings for the URL based on the screened HTML content, the one or more ratings comprising at least one rating for each of the one or more categories of content;
• storing the generated one or more ratings for the URL with an identifier for the URL in the response cache;
• sending, to an ad server, the impression request for the web page with the one or more existing ratings, in a case in which the URL has the one or more existing ratings stored in the response cache;
• sending, to the ad server, the impression request for the web page with the generated one or more ratings, in a case in which the URL has no existing ratings stored in the response cache and the number of requests for the URL has reached the threshold number of requests;
• and
• sending, to the ad server, the impression request for the web page without a rating, in a case in which the URL has no existing ratings stored in the response cache and the number of requests for the URL has not reached the threshold number of requests.