Method and system to detect invalid and fraudulent impressions and clicks in web-based advertisement systems

1. A method for detecting fraudulent activity in a pay-per-click system, comprising:

(a) providing a search engine on a server side;

b) receiving, at said server side, a search request from a client;

c) generating a unique code on the server side, for identifying said client;

d) transmitting said code to said client from the server side;

e) transmitting to said client from the server side, in response to said search request, a plurality of links associated with a plurality of websites associated with a plurality of merchants;

f) generating website information regarding a website selected by the client when said client clicks one of said plurality of links;

g) transmitting said code and said website information together from said client to said server side;

h) receiving said code and said website information at the server side, and detecting fraudulent activity by measuring the duration between clicks by said client to said selected website by examining said code and website information.

2. A method for detecting fraudulent activity in a pay-per-click system as in claim 1 further comprising the step of:

storing said code at said client.

3. A method for detecting fraudulent activity in a pay-per-click system as in claim 1 further comprising the step of:

storing said code in a cookie.

4. A method for detecting fraudulent activity in a pay-per-click system as in claim 3 wherein said storing step is performed using at least one of the following: active, Java, and Javascript.

5. A method for detecting fraudulent activity in a pay-per-click system as in claim 1 further comprising the step of:

concatenating said code with at least one of said links to said websites.

6. A method for detecting fraudulent activity in a pay-per-click system as in claim 1 wherein said code is based on a global unique identifier of said client.

7. A method for detecting fraudulent activity in a pay-per-click system as in claim 1 wherein said code is based on a network cartography of said client.

8. A method for detecting fraudulent activity in a pay-per-click system as in claim 1 wherein said code is a serial number.

9. A method for detecting fraudulent activity in a pay-per-click system as in claim 1 further comprising the step of:

preventing a click-through to at least one of said websites associated with merchants if fraudulent activity is detected.

10. In an advertising system including a search engine on a server side, the search engine providing a plurality of links associated with a plurality of web pages to a user at a device on a client side, the method for identifying fraud comprising the steps of:

generating a code on the server side, the code identifying said device on the client side;

sending said code to said device;

receiving data from said device, said data including said code and information about one or more selections by the user of at least one of said plurality of web pages;

determining from said data whether said at least one of said selections of said at least one web page is fraudulent; and

examining a duration between a time of one of said selections of said at least one web site and a time of another of said selections of said at least one web site.

11. In an advertising system including a search engine on a server side, the search engine providing a plurality of links associated with a plurality of web pages to a user at a device on a client side, the method for identifying fraud as in claim 10 wherein said code is based on a global unique identifier of said device.

12. In an advertising system including a search engine on a server side, the search engine providing a plurality of links associated with a plurality of web pages to a user at a device on a client side, the method for identifying fraud as in claim 10 wherein said code is based on a network cartography of said client.

13. In an advertising system including a search engine on a server side, the search engine providing a plurality of links associated with a plurality of web pages to a user at a device on a client side, the method for identifying fraud as in claim 10 wherein said code is a serial number.

14. A method for detecting fraud in an advertising system including a search engine on a server side comprising the steps of:

generating a code on the server side, the code identifying a device on a client side;

generating on the server side, a plurality of links associated with a plurality of web pages in response to a search initiated at the device on the client side wherein at least one of said links includes said code;

sending said plurality of links to the device;

receiving data from said device, said data including said code and information about one or more clicks by the user on at least one of said plurality of links associated with the plurality of web pages;

determining from said data whether said at least one of said clicks on said at least one link is fraudulent; and

examining a duration between a time of one of said clicks on said at least one link and a time of another of said clicks on said at least one link.

15. A method for detecting fraud in an advertising system including a search engine on a server side as in claim 14 wherein said code is based on a global unique identifier of said device.

16. A method for detecting fraud in an advertising system including a search engine on a server side as in claim 14 wherein said code is based on a network cartography of said client.

17. A method for detecting fraud in an advertising system including a search engine on a server side as in claim 14 wherein said code is a serial number.

18. A method for detecting fraudulent activity in a pay-per-click system, comprising:

(a) providing a search engine on a server side;

b) receiving, at said server side, a search request from a client;

c) generating a unique code on the server side, for identifying said client;

d) transmitting said code to said client from the server side;

e) transmitting to said client from the server side, in response to said search request, a plurality of links associated with a plurality of websites associated with a plurality of merchants;

f) receiving, on said server side, website information regarding a website selected by the client together with said code, when said client clicks one of said plurality of links;

g) detecting fraudulent activity by measuring the duration between clicks by said client to said selected website by examining said code and website information.

19. A method for detecting fraudulent activity in a pay-per-click system as in claim 18 further comprising the step of:

storing said code at said client.

20. A method for detecting fraudulent activity in a pay-per-click system as in claim 19 wherein said storing step is performed using at least one of the following: active, Java, and Javascript.

21. A method for detecting fraudulent activity in a pay-per-click system as in claim 18 further comprising the step of:

storing said code in a cookie.

22. A method for detecting fraudulent activity in a pay-per-click system as in claim 18 further comprising the step of:

concatenating said code with at least one of said links to said websites.

23. A method for detecting fraudulent activity in a pay-per-click system as in claim 18 wherein said code is based on a global unique identifier of said client.

24. A method for detecting fraudulent activity in a pay-per-click system as in claim 18 wherein said code is based on a network cartography of said client.

25. A method for detecting fraudulent activity in a pay-per-click system as in claim 18 wherein said code is a serial number.

26. A method for detecting fraudulent activity in a pay-per-click system as in claim 18 further comprising the step of:

preventing a click-through to at least one of said websites associated with merchants if fraudulent activity is detected.