SYSTEM FOR DETERMINING THE QUALITY OF QUERY SUGGESTION SYSTEMS USING A NETWORK OF USERS AND ADVERTISERS

• US 20080256039A1
• Filed: 07/30/2007
• Published: 10/16/2008
• Est. Priority Date: 04/10/2007
• Status: Abandoned Application

• Current Assignee: Oath Inc. (Verizon Communications Inc.)
• Original Assignee: Yahoo! Inc. (Apollo Global Management, Inc.)
• Inventors: Tayal, Manish, Anastasakos, Tasos, Chang, Chi-Chao
• Application Number:
  US11/830,693
• Publication Number:
  US 20080256039A1
• Time in Patent Office: Days
• US Class Current: 707/3
• CPC Class Codes: G06Q 30/02 Marketing; Price estimation or determination; Fundraising
• CPC Class Codes: G06Q 30/0244 Optimization
• CPC Class Codes: G06Q 30/0256 User search
• CPC Class Codes: G06Q 30/0277 Online advertisement

Abstract

A system is described for determining the quality of query suggestion systems using a network of users and advertisers. The system may include a memory and a processor. The memory may store data representing a plurality of query suggestion systems, a historical dataset, a plurality of residual values, and a data representing a network. The network may comprise a plurality of query items representing queries linked to a plurality of advertisement items representing advertisements via a plurality of query-advertisement link items. The processor may retrieve the historical dataset and generate the data representing the network. The processor may calculate the residual values for each query suggestion system in the plurality of query suggestion systems. The residual value may represent a value attributable to the query suggestions generated by a given query suggestion system. The processor may store the residual value of each query suggestion in the memory.

Claims

1. A method for determining quality of query suggestion systems using a network of users and advertisers, the method comprising:

identifying a plurality of query suggestion systems wherein each query suggestion system generates a plurality of query suggestions in a historical dataset;

generating, from the historical dataset, a data representing a network comprising a plurality of query items representing queries linked to a plurality of advertisement items representing advertisements via a plurality of query-advertisement link items, wherein each query-advertisement link item comprises a weight representing the strength of the association between each linked query item and each linked advertisement item, an advertisement value representing a value provided by the linked advertisement item, and a match type describing a query suggestion system in the plurality of query suggestion systems associated with the query-advertisement link item;

calculating a residual value for each query suggestion system in the plurality of query suggestion systems based on the weight, the advertisement value, and the match type each query-advertisement link item in the plurality of query-advertisement link items, wherein the residual value represents a value attributable to the plurality of query suggestions generated by each query suggestion system; and

storing the residual value for each query suggestion system in a data store.

2. The method of claim 1 wherein each query item in the plurality of query items comprises a search query performed by a user in a plurality of users.

3. The method of claim 1 wherein each advertisement item in the plurality of advertisement items comprises an online advertisement of an advertiser.

4. The method of claim 3 further comprising linking an advertisement item to a query item, via a query-advertisement link item, when the historical dataset indicates a user clicked on the advertisement item after searching for the query item.

5. The method of claim 4 wherein the match type of each query-advertisement link item indicates the query suggestion system used to generate a query suggestion for the search query which resulted in the online advertisement being displayed to the user.

6. The method of claim 5 wherein the match type of at least one query-advertisement link item comprises an exact match, further wherein the exact match indicates the online advertisement was displayed to the user as a result of the search query.

7. The method of claim 4 further comprising calculating the weight of each query-advertisement link item based on a total number of click throughs attributable to each query-advertisement link item.

8. The method of claim 7 wherein the total number of click throughs represents the number of times each user in the plurality of users clicked on the advertisement item associated with the query-advertisement link item after searching for the query item associated with the query-advertisement link item.

9. The method of claim 7 wherein the advertisement value of each query-advertisement link represents an amount paid by the advertiser generator for each click by the user on the online advertisement.

10. A method for prioritizing query suggestion systems in a data serving system, the method comprising:

identifying a data serving system wherein a plurality of query suggestion systems are used to generate a plurality of query suggestions;

generating, from a historical dataset of the data serving system, a data structure comprising a plurality of query items representing queries linked to a plurality of advertisement items representing advertisements via a plurality of query-advertisement link items wherein each query-advertisement link item is associated with a match type data describing a query suggestion system in the plurality of query suggestion systems;

calculating a residual value of each query suggestion system in the plurality of query suggestion systems based on the match type data in the data structure;

ranking the plurality of query suggestion systems based on the residual value attributed to each query suggestion system in the plurality of query suggestion systems;

decreasing the utilization of query suggestions suggested by the lowest ranking query suggestion systems in the data serving system; and

increasing the utilization of query suggestions suggested by the highest ranking query suggestion systems in the data serving system.

11. The method of claim 10 further comprising linking an advertisement item to a query item when the historical dataset indicates a user clicked on the advertisement item after searching for the query item.

12. The method of claim 10 wherein the data serving system comprises an advertisement serving system.

13. The method of claim 10 wherein decreasing the utilization of query suggestions generated by the lowest ranking query suggestion systems further comprises removing the lowest ranked query suggestion system from the data serving system.

14. The method of claim 10 further comprising calculating a weight value for each query-advertisement link item in the plurality of query-advertisement link items, wherein the weight value indicates the strength of the relationship between the query item and the advertisement item linked by each query-advertisement link item.

15. The method of claim 14 wherein the calculated weight of each query-advertisement link item in the plurality of query-advertisement link items comprises a total number of click throughs attributable to each query-advertisement link item.

16. The method of claim 15 wherein the residual value of each query suggestion system is calculated using the calculated weight of each advertisement link item associated with a match type data describing the query suggestion system.

17. A system for determining quality of query suggestion systems using a network of users and advertisers, the system comprising:

a memory to store a data representing a plurality of query suggestion systems, a historical dataset, a data representing a network comprising a plurality of query items representing queries linked to a plurality of advertisement items representing advertisements via a plurality of query-advertisement link items, wherein each query-advertisement link item comprises a weight representing the strength of the relationship between each linked query item and each linked advertisement item, an advertisement value representing a value provided by the linked advertisement item, and a match type describing the query suggestion system in the plurality of query suggestion systems associated with the query-advertisement link item, and a plurality of residual values; and

a processor operatively connected to the memory, the processor operative to retrieve the historical dataset, generate the data representing the network, calculate the residual value for each query suggestion system in the plurality of query suggestion systems based on the weight, advertisement value and match type of each query-advertisement link item in the plurality of query-advertisement link items, wherein the residual value represents a value attributable to the plurality of query suggestions generated by each query suggestion system, and storing the residual value of each query suggestion system in the memory.

18. The system of claim 17 further comprising an interface operatively connected to the memory and the processor, the interface for communicating with a user.

19. The system of claim 18 wherein the processor appends the residual value of each query suggestion system and the data describing each query suggestion system to a web page and serves the web page to the user via the interface.

20. The system of claim 17 wherein each advertisement item in the plurality of advertisement items comprises an online advertisement of an advertiser.

21. The system of claim 17 wherein the processor calculates the residual value for each query suggestion system based on the weight and the advertisement value of each query-advertisement link item associated with the match type describing the query suggestion system.