System and method for detecting spammers in a network environment

• US 9,350,636 B2
• Filed: 10/08/2013
• Issued: 05/24/2016
• Est. Priority Date: 10/08/2013
• Status: Active Grant

• Current Assignee: Match Group, LLC (Match Group, Inc.)
• Original Assignee: Match.com LLC (Match Group, Inc.)
• Inventors: Quisel, Tom R., Wang, Anson
• Primary Examiner(s): Won, Michael
• Assistant Examiner(s): Garcia-Ching, Karina J
• Application Number:
  US14/048,959
• Publication Number:
  US 20150100678A1
• Time in Patent Office: 959 Days
• Field of Search: 709/206, 709/224
• US Class Current: 1/1
• CPC Class Codes: H04L 43/10 Active monitoring, e.g. heartbeat, ping or trace-route
• CPC Class Codes: H04L 51/212 using filtering or selective blocking

Abstract

A method is provided in one example embodiment and includes processing a first text created by a user into a first bag of words, the first bag of words comprising a list of words that appear in the text, each of the words having associated therewith a number representing a number of times the associated word appears in the text; and computing a similarity between the first bag of words and at least one second bag of words. The method further comprises comparing the computed similarity with a threshold; and_determining that the user is a spammer if the computed similarity bears a first relationship with the threshold.

Claims

1. A method comprising:

processing a first text created by a user using an online service into a first bag of words, the first bag of words comprising a list of words that appear in the first text, each of the words having associated therewith a number representing a number of times the associated word appears in the text;

computing a similarity between the first bag of words and at least one second bag of words, wherein the computing comprises, for each word in the first bag of words, determining a compare count comprising a minimum number of times the word appears in each of the first bag of words and the second bag of words and adding the compare count to a sum of counts, wherein the computed similarity comprises two times the sum of counts divided by the total number of words in the first bag of words and the second bag of words;

comparing the computed similarity with a threshold; and

determining that the user is a spammer and preventing the user from using the online service to create additional texts if the computed similarity is greater than the threshold,

wherein the first text comprises a user profile of the user in connection with the online service.

2. The method of claim 1, wherein the processing a first text into a first bag of words comprises processing each of a plurality of first texts into a first bag of words.

3. The method of claim 2, wherein the second bag of words comprises a plurality of second bags of words and the computing a similarity between the first bag of words and at least one second bag of words comprises computing a similarity between each of the first bags of words and a most similar one of the second bags of words.

4. The method of claim 3 further comprising computing an average of the similarities.

5. The method of claim 4, wherein the comparing the computed similarity with a threshold comprises comparing the computed average similarity with the threshold.

6. The method of claim 1, wherein the second bag of words is associated with a spam text.

7. Logic encoded in one or more non-transitory tangible media that includes code for execution and when executed by a processor is operable to perform operations comprising:

processing a first text created by a user using an online service into a first bag of words, the first bag of words comprising a list of words that appear in the text, each of the words having associated therewith a number representing a number of times the associated word appears in the text;

computing a similarity between the first bag of words and at least one second bag of words, wherein the computing comprises, for each word in the first bag of words, determining a compare count comprising a minimum number of times the word appears in each of the first bag of words and the second bag of words and adding the compare count to a sum of counts, wherein the computed similarity comprises two times the sum of counts divided by the total number of words in the first bag of words and the second bag of words;

comparing the computed similarity with a threshold; and

determining that the user is a spammer and preventing the user from using the online service to create additional texts if the computed similarity is greater than the threshold,

wherein the first text comprises a user profile of the user in connection with the online service.

8. The logic of claim 7, wherein the processing a first text into a first bag of words comprises processing each of a plurality of first texts into a first bag of words.

9. The logic of claim 8, wherein the second bag of words comprises a plurality of second bags of words and the computing a similarity between the first bag of words and at least one second bag of words comprises computing a similarity between each of the first bags of words and a most similar one of the second bags of words.

10. The logic of claim 9 further comprising code that when executed by a processor is operable to perform operations comprising computing an average of the similarities.

11. The logic of claim 10, wherein the comparing the computed similarity with a threshold comprises comparing the computed average similarity with the threshold.

12. The logic of claim 7, wherein the second bag of words is associated with a spam text.

13. An apparatus, comprising:

a server that includes a processor and a memory, wherein the apparatus is configured to;

process a first text created by a user using an online service into a first bag of words, the first bag of words comprising a list of words that appear in the text, each of the words having associated therewith a number representing a number of times the associated word appears in the text;

compute a similarity between the first bag of words and at least one second bag of words, wherein the computing comprises, for each word in the first bag of words, determining a compare count comprising a minimum number of times the word appears in each of the first bag of words and the second bag of words and adding the compare count to a sum of counts, wherein the computed similarity comprises two times the sum of counts divided by the total number of words in the first bag of words and the second bag of words;

compare the computed similarity with a threshold; and

determine that the user is a spammer and preventing the user from using the online service to create additional texts if the computed similarity is greater than the threshold,

wherein the first text comprises a user profile of the user in connection with the online service.

14. The apparatus of claim 13, wherein the processing a first text into a first bag of words comprises processing each of a plurality of first texts into a first bag of words and the second bag of words comprises a plurality of second bags of words and wherein the computing a similarity between the first bag of words and at least one second bag of words comprises computing a similarity between each of the first bags of words and a most similar one of the second bags of words.

15. The apparatus of claim 14 further comprising computing an average of the similarities.

16. The apparatus of claim 15, wherein the comparing the computed similarity with a threshold comprises comparing the computed average similarity with the threshold.

17. The apparatus of claim 13, wherein the second bag of words is associated with a spam text.