System and method for managing a trusted email datastore

• US 7,155,738 B2
• Filed: 07/08/2004
• Issued: 12/26/2006
• Est. Priority Date: 11/04/2003
• Status: Active Grant

• Current Assignee: R2 Solutions LLC (Acacia Research Corporation)
• Original Assignee: Yahoo! Inc. (Apollo Global Management, Inc.)
• Inventors: Zhu, Jing, Hu, George
• Primary Examiner(s): Moazzami, Nasser
• Assistant Examiner(s): Parthasarathy, Pramila
• Application Number:
  US10/886,891
• Publication Number:
  US 20050097319A1
• Time in Patent Office: 901 Days
• Field of Search: 726/4, 726/18, 726/14, 713/154, 709/206
• US Class Current: 726/4
• CPC Class Codes: G06Q 10/00 Administration; Management
• CPC Class Codes: H04L 51/212 using filtering or selective blocking
• CPC Class Codes: H04L 51/48 Message addressing, e.g. address format or anonymous messages, aliases

Abstract

A Trust Email Datastore (TED) system is employed to maintain a list of message addresses with associated trust ratings. The trust rating of a message address is derived from an underlying message address relationship network around the message address of interest through various network related activities such as message sending, forwarding, deleting, blocking, marking as is/is not spam, saving to address book, etc. There may be at least two components of the trust rating between two message addresses. One component of the trust rating, called a relationship trust, may be determined based on proximity of two message addresses in a message address relationship graph. Another component of the trust rating is substantially independent of the positions of two message addresses in the message address relationship graph, and is referred to as a universal trust rating. The trust rating components can be combined and employed for spam filtering.

Claims

1. A method for communicating a message over a network, comprising:

receiving the message from a sending message address;

determining a message address relationship trust between the sending message address and a destination message address for the received message;

determining a universal trust rating for the sending message address, wherein the universal trust rating is a property of the sending message address, and not a property of the destination message address; and

if at least a combination of the message address relationship trust and the Universal trust rating indicate that the received message is trustable, providing the message to an inbox associated with the destination message address.

2. The method of claim 1, wherein the combination further comprises at least one of a sum, and a weighted sum of the message address relationship trust and the universal trust rating.

3. The method of claim 1, wherein determining the universal trust rating further comprises modifying the universal trust rating for the sending message address based, in part, on another universal trust rating associated with the destination message address.

4. The method of claim 1, wherein determining the message address relationship trust further comprises determining a degree of separation between the sending message address and the destination message address.

5. The method of claim 4, wherein determining the degree of separation further comprises employing an address relationship graph.

6. The method of claim 1, wherein determining the message address relationship trust further comprises employing a relationship strength between the sending message address and the destination message address.

7. The method of claim 1, wherein if at least a combination further comprises:

if at least the combination is greater than an upper threshold, indicating that the received message is trustable; and

if at least the combination is less than a lower threshold, indicating that the received message is untrustable.

8. The method of claim 1, wherein determining the universal trust rating further comprises:

determining another universal trust rating associated with the destination message address;

determining another message address relationship trust between the destination message address and the sending message address; and

employing the other universal trust rating and other message address relationship trust to determine the universal trust rating for the sending message address.

9. The method of claim 1, wherein determining if at least the combination is trustable further comprises determining if at least the combination is above a threshold.

10. The method of claim 1, further comprising:

if at least the combination is determined to be untrustable, performing a response that includes at least one of deleting the message, forwarding the message to a predetermined folder, and rejecting the message.

11. The method of claim 1, further comprising:

if at least the combination is determined to be between an upper threshold and a lower threshold, inclusive, performing a spam detection on at least a part of a content of the message.

12. The method of claim 1, wherein determining the universal trust rating further comprises determining the universal trust rating based, in part, on an online activity associated with at least one of sending a message, forwarding a message, deleting a message, blocking a message, responding to a message, marking a message as spam, marking a message as not spam, and saving a message address.

27. The method of claim 1, wherein determining the universal trust rating further comprises:

determining a message domain associated with the sending message address;

determining a domain universal trust rating associated with the message domain; and

employing the domain universal trust rating as the universal trust rating for the sending message address.

28. The method of claim 1, wherein determining the universal trust rating further comprises determining the universal trust rating for the sending message address based on a weighted average of universal trust ratings for message addresses within a community of message addresses.

29. The method of claim 1, further comprising:

determining universal trust ratings for each message address within a community of message address, wherein the universal trust ratings for the community evolves to a self-consistent state over time.

30. The method of claim 29, wherein the self-consistent state further comprises the universal trust ratings of each message address within the community migrating into one of three clusters consisting of a good cluster, a bad cluster, and an unknown cluster of universal trust rating values.

13. A server for communicating a message over a network, comprising:

a datastore that is configured to store trust information; and

a trust manager, coupled to the datastore, that is configured to perform actions, including;

receiving the message from a sending message address;

determining a message address relationship trust between the sending message address and a destination message address, in part, from the stored trust information;

determining a universal trust for the sending message address, wherein the universal trust rating is a property of the sending message address, and not a property of the destination message address; and

if at least a combination of the message address relationship trust and the universal trust rating indicate that the received message is trustable, providing the message to an inbox associated with the destination message address.

14. The server of claim 13, wherein the combination further comprises a sum of the message address relationship trust and the universal trust rating.

15. The server of claim 13, wherein determining the universal trust rating further comprises modifying the universal trust rating for the sending message address based, in part, on another universal trust rating associated with the destination message address.

16. The server of claim 13, wherein determining the message address relationship trust further comprises determining a degree of separation between the sending message address and the destination message address.

17. The server of claim 13, wherein determining the message address relationship trust further comprises employing a relationship strength between the sending message address and the destination message address.

18. A system for communicating a message over a network, comprising:

a client device that is configured to perform actions, including;

registering a destination message address with a trust server; and

the trust server, coupled to the client device, that is configured to perform actions, including;

storing the registered destination message address in a datastore;

receiving the message from a sending message address;

determining a message address relationship trust between the sending message address and the destination message address, in part, from the datastore;

determining a universal trust rating for the sending message address, wherein the universal trust rating is a property of the sending message address, and not a property of the destination message address; and

if at least a combination of the message address relationship trust and the universal trust rating indicate that the received message is trustable, providing the message to an inbox associated with the destination message address.

19. The system of claim 18, wherein the trust server is configured to perform actions, further comprising:

if at least the combination is determined to be untrustable, performing a response that includes at least one of deleting the message, forwarding the message to a predetermined folder, and rejecting the message.

20. The system of claim 18, wherein the trust server is configured to perform actions, further comprising:

if at least the combination is determined to be between an upper threshold and a lower threshold, inclusive, performing a spam detection on at least a part of a content of the message.

21. The system of claim 18, wherein determining the universal trust rating further comprises determining the universal trust rating based, in part, on an online activity associated with at least one of sending a message, forwarding a message, deleting a message, blocking a message, responding to a message, marking a message as spam, marking a message as not spam, and saving a message address.

22. A computer-readable medium having computer-executable instructions for communicating a message over a network, the computer-executable instructions being operative to perform actions, comprising:

receiving the message from a sending message address over the network;

determining a message address relationship trust between the sending message address and a destination message address for the message;

determining a universal trust rating associated with the sending message address, wherein the universal trust rating is a property of the sending message address, and not a property of the destination message address; and

if at least a combination of the message address relationship trust and the universal trust rating indicate that the received message is trustable, providing the message to an inbox associated with the destination message address.

23. The computer-readable medium of claim 22, further comprising:

if at least the combination is determined to be untrustable, performing a response that includes at least one of deleting the message, forwarding the message to a predetermined folder, and rejecting the message.

24. The computer-readable medium of claim 22, further comprising:

if at least the combination is determined to be between an upper threshold and a lower threshold, inclusive, performing a spam detection on at least a part of a content of the message.

25. The computer-readable medium of claim 22, wherein determining the universal trust rating further comprises determining the universal trust rating based, in part, on an online activity associated with at least one of sending a message, forwarding a message, deleting a message, blocking a message, responding to a message, marking a message as spam, marking a message as not spam, and saving a message address.

26. A server for communicating a message on a network, comprising:

a means for receiving the message from a sending message address;

a means for determining a message address relationship trust between the sending message address and a destination message address for the received message;

a means for determining a universal trust rating for the sending message address, wherein the universal trust rating is a property of the sending message address, and not a property of the destination message address; and

if at least a combination means of the message address relationship trust and the universal trust relationship rating indicate that the received message is trustable, a means for providing the message to an inbox associated with the destination message address.