Real-time ad-hoc spam filtering of email

• US 8,069,128 B2
• Filed: 08/08/2008
• Issued: 11/29/2011
• Est. Priority Date: 08/08/2008
• Status: Active Grant

• Current Assignee: Yahoo Assets LLC
• Original Assignee: Yahoo! Inc. (Apollo Global Management, Inc.)
• Inventors: Pujara, Jay
• Primary Examiner(s): Gaffin; Jeffrey A
• Assistant Examiner(s): SITIRICHE, LUIS A
• Application Number:
  US12/188,612
• Publication Number:
  US 20100036786A1
• Time in Patent Office: 1,208 Days
• Field of Search: None
• US Class Current: 706/46
• CPC Class Codes: G06Q 10/107 Computer-aided management of electronic mailing [e-mailing]
• CPC Class Codes: H04L 51/212 using filtering or selective blocking

Abstract

Embodiments are directed towards employing a multi-pass ad-hoc spam message filtering approach that dynamically generates a temporary classifier during a first pass based on a result of a previously applied message filter that sorts messages into various folders for a user. The first pass scans messages in a user'"'"'s mail folders, and reads various information within the messages, including, but not limited to sender information, headers, including a subject, an originating network address, message contents, attachments, and the like. After creating a classification model, the classifier with its model is used in a second pass on the message folders to retrospectively inspect the messages and present to the user a list of messages that might be misclassified. The classification model is maintained within memory on a user'"'"'s client device, as memory resident only, and is not stored on disk or within another persistent data store.

Claims

1. A network device to manage a delivery of a message over a network, comprising:

a transceiver to send and receive data over a network; and

a processor that is operative to perform actions, comprising;

performing a first pass analysis on a plurality of messages, which have been previously sorted into a plurality of different message folders by a first message filter, using a second, multi-pass message filter that is configured to generate a classification model based on the first pass analysis, wherein the generated classification model is configured to reside exclusively in a non-persistent memory location;

using the generated classification model with the second, multi-pass message filter to perform a second pass on the sorted plurality of messages, to generate a misclassification identification of sorted messages;

displaying the misclassification identification of sorted messages such that at least one misclassified message can be moved to a different message folder based on the misclassification identification; and

purging the generated classification model from non-persistent memory location.

2. The network device of claim 1, wherein performing a first pass analysis on the sorted plurality of messages is configured to train the second, multi-pass message filter based on which folder each of the plurality of messages resides.

3. The network device of claim 1, wherein the generated classification model is generated based on an analysis of at least which message folder a message currently resides.

4. The network device of claim 1, wherein the classification model is generated based on an analysis of at least one of a state of a message, whether a message sender'"'"'s address in an address book associated with a recipient that is associated with the message folders, a header of a message, or a message content.

5. The network device of claim 1, wherein access to the generated classification model is restricted.

6. A processor readable storage medium that includes data and instructions, wherein the execution of the instructions on a computing device provides for managing messages by enabling actions, comprising:

receiving a plurality of messages;

sorting the plurality of messages into a plurality of different message folders using a first message filter, wherein the first message filter is configured to store first classification data in a persistent store;

executing a first pass of a multi-pass message filter on the sorted plurality of messages to train the multi-pass message filter by generating second classification data, wherein the second classification data is configured to reside exclusively in a non-persistent memory store;

using the second classification data, executing a second pass of the multi-pass message filter on the sorted plurality of messages to score each message;

identifying any message in the plurality of messages that based on the message'"'"'s respective score indicates that the message is sorted into an incorrect message folder; and

providing a display of information associated with any message identified as stored into the incorrect message folder, such that a user is enabled to move the identified message to another message folder.

7. The processor readable storage medium of claim 6, wherein the instructions enable actions, further comprising deleting second classification data such that it is no longer accessible.

8. The processor readable storage medium of claim 6, wherein executing the first pass and the second pass of the multi-pass message filter occurs entirely on a client device.

9. The processor readable storage medium of claim 6, wherein the first message filter is configured to execute on a network device distinct from a client device.

10. The processor readable storage medium of claim 6, wherein the multi-pass message filter employs a Naive Bayes classifier.

11. The processor readable storage medium of claim 6, wherein the generated second classification model is generated based on an analysis of at least which message folder a message currently resides.

12. A method for managing a message delivery, comprising:

receiving a plurality of messages;

sorting the plurality of messages into a plurality of different message folders using a first message filter, wherein the first message filter is configured to store first classification data in a persistent store;

executing, on a client device, a first pass of a multi-pass message filter on the sorted plurality of messages to train the multi-pass message filter by generating second classification data, wherein the second classification data is configured to reside exclusively in a non-persistent memory store;

using the second classification data, executing on the client device a second pass of the multi-pass message filter on the sorted plurality of messages to score each message;

identifying any message in the plurality of messages that based on the message'"'"'s respective score indicates that the message is sorted into an incorrect message folder; and

providing a display of information associated with any message identified as stored into the incorrect message folder, such that a user is enabled to move the identified message to another message folder.

13. The method of claim 12, wherein the second classification data is deleted from the client device upon providing the display of information.

14. The method of claim 12, wherein the multi-pass message filter is selectable for execution by a user through an interface on the client device;

and wherein the first message filter is automatically executed independent of a user action.

15. The method of claim 12, wherein the second classification data is generated based on an analysis of at least one of a state of a message, whether a message sender'"'"'s address in an address book associated with a recipient that is associated with the message folders, a header of a message, or a message content.

16. The method of claim 12, wherein the second classification data is generated based on which folder each of the plurality of messages resides.

17. A system for enabling a communications over a network, comprising:

a network device configured to perform actions, including;

receiving a plurality of messages;

assigning the plurality of messages to a plurality of different message folders using a first message filter, wherein the first message filter is configured to store first classification data in a persistent store; and

communicating the messages over the network to the plurality of different message folders; and

a client device configured to perform actions, including;

executing a first pass of a multi-pass message filter on the sorted plurality of messages to train the multi-pass message filter by generating second classification data, wherein the second classification data is configured to reside exclusively in a non-persistent memory store at the client device;

using the second classification data, executing a second pass of the multi-pass message filter on the sorted plurality of messages to score each message;

identifying any message in the plurality of messages that based on the message'"'"'s respective score indicates that the message is sorted into an incorrect message folder; and

providing a display of information associated with any message identified as stored into the incorrect message folder, such that a user is enabled to move the identified message to another message folder.

18. The system of claim 17, wherein the second classification data is deleted from the client device upon providing the display of information.

19. The system of claim 17, wherein execution of the multi-pass message filter is user selectable.

20. The system of claim 17, wherein the second classification data is generated based on an analysis of at least one of a state of a message, whether a message sender'"'"'s address in an address book associated with a recipient that is associated with the message folders, a header of a message, or a message content.