System and method for spam identification

• US 7,660,792 B2
• Filed: 04/29/2005
• Issued: 02/09/2010
• Est. Priority Date: 04/29/2005
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Corporation
• Inventors: Watson, Eric B., Brewer, Brett D.
• Primary Examiner(s): Alam; Hosain T
• Assistant Examiner(s): Ng; Amy
• Application Number:
  US11/117,568
• Publication Number:
  US 20060248072A1
• Time in Patent Office: 1,747 Days
• Field of Search: 707/3, 707/4, 707/5, 707/10, 705/10, 709/218, 709/203, 709/204
• US Class Current: 726/22
• CPC Class Codes: G06F 16/904 Browsing; Visualisation therefor for navigating the web G06F16/954; browsing optimisation for the web G06F16/957
• CPC Class Codes: G06F 16/954 Navigation, e.g. using categorised browsing

Abstract

A system and method are provided for improving a user search experience by identifying spam results in a result set produced in response to a query. The system may include a user interface spam feedback mechanism for allowing a user to indicate that a given result is spam. The system may additionally include an automated spam identification mechanism for implementing automated techniques on the given result to determine whether the given result is spam. The system may further include a merging component for merging the determinations of the user interface spam feedback mechanism and the automated spam identification mechanism for deriving an indicator of the likelihood that a given result is spam.

Claims

1. A method for improving a user search experience by identifying spam results in a result set produced in response to a query, the method comprising:

receiving, at a computing device, user feedback identifying as spam an individual result within a plurality of search results presented to a user in response to a query entered by the user, wherein the user feedback is entered by selecting a link that is displayed contemporaneously with and adjacent to the individual result;

implementing one or more automated spam identification techniques on the individual result to determine whether the individual result is spam based on one or more characteristics of a website referenced by the individual result; and

merging data obtained from the user feedback and the one or more automated spam identification techniques to obtain an indicator for the individual result, the indicator providing a likelihood that the individual result is spam, wherein the one or more automated spam identification techniques include determining a keyword monetization value for the query, wherein when the individual result is returned in response to a specific query with a low keyword monetization value the individual result is less likely to be spam.

2. The method of claim 1, further comprising analyzing user feedback received through the user interface to determine if the user feedback is spam, wherein the user feedback is designated as spam if above a threshold amount of the user feedback is originating from less than a threshold amount of IP addresses.

3. The method of claim 1, wherein the one or more automated spam identification techniques comprise determining whether keyword stuffing is present on the website referenced by the individual result.

4. The method of claim 1, further comprising analyzing the user feedback for the individual result across multiple queries, wherein a confidence that the individual result is spam is increased when the user feedback is received in connection with presenting the individual result in response to the multiple queries.

5. The method of claim 1, wherein the one or more automated spam identification techniques comprise determining a query independent rank of the individual result.

6. The method of claim 1, wherein the one or more automated spam identification techniques comprise a popularity analysis that determines a popularity of the individual result by examining traffic to a website referenced by the individual result, wherein a high popularity indicates a lower probability that the individual result is spam.

7. The method of claim 1, further comprising storing the individual result and the indicator in an index in order so that a ranking mechanism accesses the indicator when the given result is produced in response to any future query.

8. The method of claim 1, further comprising determining that the individual result is spam if the one or more automated spam identification techniques for the individual result and the user input agree that the individual result is spam.

9. A method for improving a user search experience by identifying spam results in a result set produced in response to a query, the method comprising:

receiving, at a computing device, a search query;

providing a set of search results;

providing a user interface spam feedback mechanism for allowing a user to explicitly designate an individual result within the set of search results as spam, wherein the spam feedback mechanism is displayed contemporaneously with the set of search results;

receiving user feedback designating the individual result as spam;

determining whether the user feedback is spam feedback;

implementing one or more automated spam identification techniques on the individual result to determine whether the individual result is spam; and

merging data obtained from the user feedback and the one or more automated spam identification techniques to obtain an indicator for the individual result, the indicator providing a likelihood that the individual result is spam, wherein the one or more automated spam identification techniques include determining a keyword monetization value for the query, wherein when the individual result is returned in response to a specific query with a low keyword monetization value the individual result is less likely to be spam.

10. The method of claim 9, wherein the spam feedback is disregarded when altering the future ranking.

11. The method of claim 9, wherein determining whether user feedback is spam comprises determining a source of the feedback and determining whether the feedback originates with unique users.

12. The method of claim 9, further comprising analyzing the user feedback for the individual result across multiple queries, wherein a confidence that the individual result is spam is increased when user feedback indicating the individual result is spam is received in connection with presenting the individual result in response to the multiple queries.

13. The method of claim 9, further comprising determining that the individual result is spam if the one or more automated spam identification techniques for the individual result and the user input agree that the individual result is spam.

14. The method of claim 9, wherein the one or more automated spam identification techniques comprise determining a query independent rank of the individual result.

15. The method of claim 9, wherein the one or more automated spam identification techniques comprise a popularity analysis that determines a popularity of the individual result by examining traffic to a website referenced by the individual result, wherein a high popularity indicates a lower probability that the individual result is spam.

16. The method of claim 9, further comprising storing the individual result and the spam indicator in an index in order so that a ranking mechanism accesses the indicator when the individual result is produced in response to any future query.

17. One or more computer storage media storing computer-executable instructions embodied thereon for performing a method of identifying spam search results, the method comprising:

receiving, at a computing device, user feedback from one or more users who enter the user feedback into a user interface that is displayed designate one or more search results as spam;

determining whether the user feedback is spam;

aggregating the user feedback for each of the one or more search results into a feedback analysis;

implementing one or more automated spam identification techniques on each of the one or more search results to generate an automated analysis that determines whether each search result is spam; and

merging the feedback analysis and the automated analysis for each of the one or more search results to generate an indicator for each of the one or more search results, wherein the indicator indicates a likelihood that the search result is spam, wherein the one or more automated spam identification techniques include determining a keyword monetization value for the query, wherein when the individual result is returned in response to a specific query with a low keyword monetization value the individual result is less likely to be spam.

18. The one or more media of claim 17, wherein the user feedback is designated as spam if above a threshold amount of the user feedback originated from less than a threshold amount of IP addresses.