CLUSTERING COMMUNICATIONS BASED ON CLASSIFICATION

• US 20160314182A1
• Filed: 09/18/2014
• Published: 10/27/2016
• Est. Priority Date: 09/18/2014
• Status: Active Grant

• Current Assignee: Google LLC (Alphabet Inc.)
• Original Assignee: Google Inc. (Alphabet Inc.)
• Inventors: Zhang, Xincheng, Tan, Hui, Wang, Zhiyu, Lou, Jinan
• Granted Patent: US 10,007,717 B2
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: G06F 16/285 Clustering or classification
• CPC Class Codes: G06F 16/35 Clustering; Classification
• CPC Class Codes: H04L 43/04 Processing captured monitoring data, e.g. for logfile generation

Abstract

Methods and apparatus related to clustering documents based on one or more classification terms and optionally based on similarity of structural paths of the documents. In some implementations, the documents are communications such as structured emails or other structured communications. In some of those implementations, clustering the communications includes identifying a plurality of classification terms indicative of a classification, identifying a corpus of communications that includes communications that are not labeled with an association to the classification, and determining a cluster of the communications based on occurrence of one or more of the classification terms in the communications of the cluster.

Claims

1. A computer implemented method, comprising:

identifying a plurality of classification terms indicative of a classification;

identifying a corpus of communications from one or more databases, the corpus of communications including a plurality of communications that are not labeled with an association to the classification;

determining a cluster of the communications based on occurrence of one or more of the classification terms in the communications of the cluster;

determining a feature set based on the communications of the cluster, the feature set including one or more features that are based on content of the communications that is in addition to the classification terms; and

assigning the feature set to an indication of the classification.

2. The method of claim 1, wherein the communications of the cluster include structured communications, and wherein determining the cluster of the communications is further based on structural paths of the structured communications of the cluster.

3. The method of claim 2, wherein determining the feature set comprises:

determining structural features of the features based on the structural paths of the structured communications of the cluster.

4. The method of claim 3, wherein determining the feature set comprises:

determining term features of the features based on communication terms of the structured communications of the cluster;

wherein one or more of the communication terms on which the term features are determined are in addition to the classification terms.

5. The method of claim 1, wherein determining the feature set comprises:

determining term features of the features based on communication terms of the communications of the cluster;

wherein one or more of the communication terms on which the term features are determined are in addition to the classification terms.

6. The method of claim 1, wherein the communications of the cluster include structured communications, and wherein determining the feature set comprises:

determining term features of the features based on communication terms of the structured communications of the cluster and structural paths of the structured communications of the cluster, wherein each of the term features is indicative of a term of the communication terms and one or more of the structural paths in which the term appears in the structured communications of the given cluster.

7. The method of claim 1, wherein the communications of the cluster are from a plurality of sending entities.

8. The method of claim 1, wherein identifying the classification terms indicative of the classification comprises:

identifying one or more labeled communications having labels indicating the classification; and

determining one or more of the classification terms based on the labeled communications.

9. The method of claim 8, wherein the labels indicating the classification are based on names of folders with which the labeled communications are associated.

10. The method of claim 8, wherein the labels indicating the classification are based on annotations of the labeled communications by users that received the labeled communications.

11. The method of claim 8, wherein determining the one or more of the classification terms based on the labeled communications comprises:

selecting a term of the classification terms based on comparison of a frequency of occurrence of the term in the labeled communications to a frequency of occurrence of the term in a larger corpus of documents.

12. The method of claim 1, wherein the communications are emails.

13. The method of claim 1, further comprising:

identifying an additional communication; and

determining the additional communication is associated with the indication of the classification based on comparing the feature set to content of the additional communication.

14. The method of claim 1, wherein the indication of the classification is a data extraction parser for that classification.

15. The method of claim 14, wherein the data extraction parser is configured based on the communications of the given cluster.

16. A system including memory and one or more processors operable to execute instructions stored in the memory, comprising instructions to:

identify a plurality of classification terms indicative of a classification;

identify a corpus of communications from one or more databases, the corpus of communications including a plurality of communications that are not labeled with an association to the classification;

determine a cluster of the communications based on occurrence of one or more of the classification terms in the communications of the cluster;

determine a feature set based on the communications of the cluster, the feature set including one or more features that are based on content of the communications that is in addition to the classification terms; and

assign the feature set to an indication of the classification.

17. The system of claim 16, wherein the communications of the cluster include structured communications, and wherein the instructions to determine the cluster of the communications further include instructions to determine the cluster based on structural paths of the structured communications of the cluster.

18. The system of claim 17, wherein the instructions to determine the feature set comprise instructions to:

determine structural features of the features based on the structural paths of the structured communications of the cluster.

19. The system of claim 18, wherein the instructions to determine the feature set comprise instructions to:

determine term features of the features based on communication terms of the structured communications of the cluster;

wherein one or more of the communication terms on which the term features are determined are in addition to the classification terms.

20. The system of claim 16, wherein the instructions to determine the feature set comprise instructions to:

determine term features of the features based on communication terms of the communications of the cluster;

wherein one or more of the communication terms on which the term features are determined are in addition to the classification terms.

21. The system of claim 16, wherein the communications of the cluster include structured communications, and wherein the instructions to determine the feature set comprise instructions to:

determine term features of the features based on communication terms of the structured communications of the cluster and structural paths of the structured communications of the cluster, wherein each of the term features is indicative of a term of the communication terms and one or more of the structural paths in which the term appears in the structured communications of the given cluster.

22. The system of claim 16, wherein the instructions to identify the classification terms indicative of the classification comprise instructions to:

identify one or more labeled communications having labels indicating the classification; and

determine one or more of the classification terms based on the labeled communications.

23. The system of claim 22, wherein the labels indicating the classification are based on names of folders with which the labeled communications are associated.

24. The system of claim 22, wherein the labels indicating the classification are based on annotations of the labeled communications by users that received the labeled communications.

25. The system of claim 16, wherein the instructions further comprise instructions to:

identify an additional communication; and

determine the additional communication is associated with the indication of the classification based on comparing the feature set to content of the additional communication.

26. A non-transitory computer-readable storage medium comprising instructions that, in response to execution of the instructions by a computing system, cause the computing system to perform operations comprising:

identifying a plurality of classification terms indicative of a classification;

identifying a corpus of communications from one or more databases, the corpus of communications including a plurality of communications that are not labeled with an association to the classification;

determining a cluster of the communications based on occurrence of one or more of the classification terms in the communications of the cluster;

determining a feature set based on the communications of the cluster, the feature set including one or more features that are based on content of the communications that is in addition to the classification terms; and

assigning the feature set to an indication of the classification.