Preservation of scores of the quality of traffic to network sites across clients and over time
A software and/or hardware facility for scoring the quality of traffic to a site accessible via the Internet or other network. The facility may generate training set data and use the training set data to identify parameters indicative of fraudulent traffic to a site and reduce the effect of fraudulent traffic advertisers and publishers. The facility may score the quality of traffic and determine combinations of parameters that are indicative of the quality of traffic to the site. Traffic to the site may be scored based on the combination of parameters associated with the one or more sessions. Lower scores are indicative of traffic having little value to a publisher, advertiser, or third party; higher scores are indicative of traffic having greater value.
1. A method comprising:
receiving, by a computer system comprising one or more hardware processors, a first set of data that identifies a first plurality of clicks by a set of devices when the devices interact via a network environment with an advertisement displayed on a publisher website during a first time period, and a second set of data that identifies a second plurality of clicks by the set of devices when the devices interact via the network environment with the advertisement displayed on the publisher website during a second time period, wherein each click of the first plurality of clicks and the second plurality of clicks is associated with a combined identifier that includes: (i) a device identifier of a device of the set of devices; and (ii) at least two bytes of an IP address associated with the device identifier;
assigning, by the computer system, values to the first plurality of clicks and the second plurality of clicks, wherein the values indicate quality of the clicks, and wherein assigning values to the clicks comprises, for each combined identifier:
measuring a velocity metric between at least two consecutive clicks associated with the combined identifier,
determining an elapsed time during which the device associated with the combined identifier interacted with the advertisement, and
assigning the values to clicks associated with the combined identifier based at least in part on the measured velocity metric and the elapsed time;
determining, by the computing system, that a first value assigned to the first plurality of clicks is higher than a second value assigned to the second plurality of clicks;
determining, by the computer system, that one or more first parameters associated with the first plurality of clicks overlap with one or more second parameters associated with the second plurality of clicks; and
in response to the determining that the one or more first parameters overlap with the one or more second parameters, modifying, by the computer system, the second value in accordance with the first value.
2. The method of claim 1 , wherein the first set of data and the second set of data are received from a server log.
3. The method of claim 1 , wherein assigning a value to the clicks is further based on IP addresses associated with the clicks.
4. The method of claim 1 , wherein the first plurality of clicks represents a first set of network sessions between the set of devices and the publisher website, and the second plurality of clicks represents a second set of network sessions between the set of devices and the publisher website.
5. The method of claim 4 , wherein determining the first value assigned to the first plurality of clicks comprises:
processing, by the computer system, the first set of data to generate a set of result vectors, wherein each result vector of the set of result vectors includes a value identifying a probability of whether a network session of the first set of network sessions includes one or more lower quality clicks; and
determining, by the computer system, the first score based on an average of the values of the set of result vectors.
6. The method of claim 1 , wherein higher quality clicks are indicative of network traffic having desirable characteristics.
7. The method of claim 1 , wherein lower quality clicks are indicative of fraudulent, likely fraudulent, or otherwise non-productive network traffic.
8. A system comprising:
one or more data processors; and
a non-transitory computer readable storage medium containing instructions which, when executed on the one or more data processors, cause the one or more data processors to perform one or more operations including:
receiving a first set of data that identifies a first plurality of clicks by a set of devices when the devices interact via a network environment with an advertisement displayed on a publisher website during a first time period, and a second set of data that identifies a second plurality of clicks by the set of devices when the devices interact via the network environment with the advertisement displayed on the publisher website during a second time period, wherein each click of the first plurality of clicks and the second plurality of clicks is associated with a combined identifier that includes: (i) a device identifier of a device of the set of devices; and (ii) at least two bytes of an IP address associated with the device identifier;
assigning values to the first plurality of clicks and the second plurality of clicks, wherein the values indicate quality of the clicks, and wherein assigning values to the clicks comprises, for each combined identifier:
measuring a velocity metric between at least two consecutive clicks associated with the combined identifier,
determining an elapsed time during which the device associated with the combined identifier interacted with the advertisement, and
assigning values to clicks associated with the combined identifier based at least in part on the measured velocity metric and the elapsed time;
determining that a first value assigned to the first plurality of clicks is higher than a second value assigned to the second plurality of clicks;
determining that one or more first parameters associated with the first plurality of clicks overlap with one or more second parameters associated with the second plurality of clicks; and
in response to the determining that the one or more first parameters overlap with the one or more second parameters, modifying the second value in accordance with the first value.
9. The system of claim 8 , wherein the first set of data and the second set of data are received from a server log.
10. The system of claim 8 , wherein assigning a value to the clicks is further based on IP addresses associated with the clicks.
11. The system of claim 8 , wherein the first plurality of clicks represents a first set of network sessions between the set of devices and the publisher website, and the second plurality of clicks represents a second set of network sessions between the set of devices and the publisher website.
12. The system of claim 11 , wherein determining the first value assigned to the first plurality of clicks comprises:
processing the first set of data to generate a set of result vectors, wherein each result vector of the set of result vectors includes a value identifying a probability of whether a network session of the first set of network sessions includes one or more lower quality clicks; and
determining the first score based on an average of the values of the set of result vectors.
13. The system of claim 8 , wherein higher quality clicks are indicative of network traffic having desirable characteristics.
14. The system of claim 8 , wherein lower quality clicks are indicative of fraudulent, likely fraudulent, or otherwise non-productive network traffic.
15. A computer-program product tangibly embodied in a non-transitory machine-readable storage medium, including instructions configured to cause one or more data processors of a computer system to perform one or more operations including:
receiving a first set of data that identifies a first plurality of clicks by a set of devices when the devices interact via a network environment with an advertisement displayed on a publisher website during a first time period, and a second set of data that identifies a second plurality of clicks by the set of devices when the devices interact via the network environment with the advertisement displayed on the publisher website during a second time period, wherein each click of the first plurality of clicks and the second plurality of clicks is associated with a combined identifier that includes: (i) a device identifier of a device of the set of devices; and (ii) at least two bytes of an IP address associated with the device identifier;
assigning values to the first plurality of clicks and the second plurality of clicks, wherein the values indicates quality of the clicks, and wherein assigning values to the clicks comprises, for each combined identifier:
measuring a velocity metric between at least two consecutive clicks associated with the combined identifier,
determining an elapsed time during which the device associated with the combined identifier interacted with the advertisement, and
assigning values to clicks associated with the combined identifier based at least in part on the measured velocity metric and the elapsed time;
determining that a first value assigned to the first plurality of clicks is higher than a second value assigned to the second plurality of clicks;
determining that one or more first parameters associated with the first plurality of clicks overlap with one or more second parameters associated with the second plurality of clicks; and
in response to the determining that the one or more first parameters overlap with the one or more second parameters, modifying the second value in accordance with the first value.
16. The computer-program product of claim 15 , wherein the first set of data and the second set of data are received from a server log.
17. The computer-program product of claim 15 , wherein assigning a value to the clicks is further based on IP addresses associated with the clicks.
18. The computer-program product of claim 15 , wherein the first plurality of clicks represents a first set of network sessions between the set of devices and the publisher website, and the second plurality of clicks represents a second set of network sessions between the set of devices and the publisher website.
19. The computer-program product of claim 15 , wherein determining the first value assigned to the first plurality of clicks comprises:
processing the first set of data to generate a set of result vectors, wherein each result vector of the set of result vectors includes a value identifying a probability of whether a network session of the first set of network sessions includes one or more lower quality clicks; and
determining the first score based on an average of the values of the set of result vectors.
20. The computer-program product of claim 15 , wherein higher quality clicks are indicative of network traffic having desirable characteristics and lower quality clicks are indicative of fraudulent, likely fraudulent, or otherwise non-productive network traffic.