SYSTEM AND METHOD FOR MESSAGE CLUSTERING

• US 20090177484A1
• Filed: 06/24/2008
• Published: 07/09/2009
• Est. Priority Date: 01/06/2008
• Status: Active Grant

• Current Assignee: Yahoo AD Tech LLC
• Original Assignee: Yahoo! Inc. (Apollo Global Management, Inc.)
• Inventors: Davis, Marc Eliot, O'Sullivan, Joseph James, Callan, Paul, Wroblewski, Luke
• Granted Patent: US 8,762,285 B2
• Time in Patent Office: Days
• US Class Current: 705/346
• CPC Class Codes: G06F 16/285 Clustering or classification
• CPC Class Codes: G06Q 30/02 Marketing; Price estimation or determination; Fundraising
• CPC Class Codes: G06Q 30/0281 Customer communication at a business location, e.g. providing product or service information, consulting
• CPC Class Codes: H04L 51/00 User-to-user messaging in packet-switching networks, transmitted according to store-and-forward or real-time protocols, e.g. e-mail
• CPC Class Codes: H04L 51/214 using selective forwarding
• CPC Class Codes: H04L 51/52 for supporting social networking services

Abstract

The disclosure describes systems and methods delivering communications associated with delivery conditions in which the occurrence of the delivery condition is determined by monitoring information received from a plurality of sources via multiple communication channels. The message delivery systems allow messages to be delivered to any “Who, What, When, Where” from any “Who, What, When, Where” upon the detection of an occurrence of one or more “Who, What, When, Where” delivery conditions. A message (which may be any data object including text-based messages, audio-based message such as voicemail or other audio such as music or video-based prerecorded messages) is delivered in accordance with delivery conditions based on any available data, including topical, spatial, temporal, and/or social data. Furthermore, because the systems coordinate delivery of messages via multiple communication channels and through multiple devices, the communication channel for delivery of a message may be dynamically determined based on the delivery conditions.

Claims

1. A method comprising:

receiving a plurality of messages for delivery on a network, the messages each containing an identity of at least one recipient;

parsing the received messages to extract message content;

collecting recipient data available to the network by searching for social data, spatial data, temporal data and logical data relating to the message recipients and the extracted message content;

analyzing the messages to identify relationships between the extracted message content and the message recipients;

grouping the messages into a cluster based upon the identified relationships and a rule set, the cluster comprising messages that satisfy rules of the rule set; and

analyzing the cluster to determine a cluster label for identification of the cluster to the message recipients.

2. The method of claim 1 further comprising:

transmitting the cluster to the message recipients;

providing the message recipients the ability to provide feedback via interaction with the cluster label, wherein the cluster label is viewable to the message recipients upon transmittal.

3. The method of claim 2 further comprising:

determining if feedback is supplied by the message recipients.

4. The method of claim 3 further comprising:

upon detection of feedback supplied from the message recipients, performing an action on the cluster based on the identified relationships associated with the cluster and the feedback provided by the message recipients.

5. The method of claim 3 further comprising:

upon no feedback supplied from the message recipients, performing an action on the cluster based upon the identified relationships associated with the cluster.

6. The method of claim 4, wherein the feedback supplied by the message recipients is based upon the message recipients performing at least one of clicking on or dragging the cluster label to seek additional information.

7. The method of claim 1, wherein the rule set represents a constraint resolution technique for building clusters based on message content of the messages.

8. The method of claim 7, wherein the constraint resolution technique is a degrees of relationship computation between the social data, spatial data, temporal data and logical data of the message recipients and a sender of one of the plurality of messages.

9. The method of claim 7, wherein the rule set is based upon user preferences that identify the message content that is of interest to the message recipients.

10. The method of claim 1, wherein the cluster is contained within a hierarchy of clusters.

11. The method of claim 1, wherein the cluster overlaps with at least one other cluster.

12. The method of claim 1, wherein the cluster is formed based upon a specific event.

13. The method of claim 1, wherein the cluster is formed upon a time period.

14. The method of claim 1, wherein the cluster is formed based upon a specific location or the at least one recipient.

15. The method of claim 1, wherein the cluster is formed based upon a distance between the message recipients and sender.

16. The method of claim 1 wherein the determining the cluster label further comprises:

labeling the cluster pertaining to the message content of the cluster that is understood by each of the message recipients.

17. The method of claim 16, wherein each of the message recipients have the option to alter the cluster label to a representation that is contextually meaningful for the user.

18. The method of claim 6, wherein said seeking of additional information comprises:

correlating the message cluster with at least one of the social data, spatial data, temporal data and logical data to identify refined content pertaining to at least one of the message content of the messages and other message recipients.

19. A computer-readable medium tangibly encoded with instructions to perform a method comprising:

receiving a plurality of messages for delivery on a network, the messages each containing an identity of at least one recipient;

parsing the received messages to extract message content;

collecting recipient data available to the network by searching for social data, spatial data, temporal data and logical data relating to the message recipients and the extracted message content;

analyzing the messages to identify relationships between the extracted message content and the message recipients;

grouping the messages into a cluster based upon the identified relationships and a rule set, the cluster comprising messages that satisfy rules of the rule set; and

analyzing the cluster to determine a cluster label for identification of the cluster to the message recipients.

20. The computer-readable medium of claim 19 further comprising:

transmitting the cluster to the message recipients;

providing the message recipients the ability to provide feedback via interaction with the cluster label, wherein the cluster label is viewable to the message recipients upon transmittal.

21. The computer-readable medium of claim 20 further comprising:

determining if feedback is supplied by the message recipients.

22. The computer-readable medium of claim 21 further comprising:

upon detection of feedback supplied from the message recipients, performing an action on the cluster based on the identified relationships associated with the cluster and the feedback provided by the message recipients.

23. The computer-readable medium of claim 21 further comprising:

upon no feedback supplied from the message recipients, performing an action on the cluster based upon the identified relationships associated with the cluster.

24. The computer-readable medium of claim 22, wherein the feedback supplied by the message recipients is based upon the message recipients performing at least one of clicking on or dragging the cluster label to seek additional information.

25. The computer-readable medium of claim 19, wherein the rule set represents a constraint resolution technique for building clusters based on message content of the messages.

26. The computer-readable medium of claim 25, wherein the constraint resolution technique is a degrees of relationship computation between the social data, spatial data, temporal data and logical data of the message recipients and a sender of one of the plurality of messages.

27. The computer-readable medium of claim 25, wherein the rule set is based upon user preferences that identify the message content that is of interest to the message recipients.

28. The computer-readable medium of claim 19, wherein the determining the cluster label further comprises:

labeling the cluster pertaining to the message content of the cluster that is understood by each of the message recipients.

29. The computer-readable medium of claim 28, wherein each of the message recipients have the option to alter the cluster label to a representation that is contextually meaningful for the user.

30. The computer-readable medium of claim 24, wherein said seeking of additional information comprises:

correlating the message cluster with at least one of the social data, spatial data, temporal data and logical data to identify refined content pertaining to at least one of the message content of the messages and other message recipients.

31. A computer system comprising:

a plurality of processors;

an attention engine implemented on one of the plurality of processors for receiving a plurality of messages for at least one recipient via a network, wherein the messages contain an identity of at least one recipient;

an attribution engine implemented on one of the plurality of processors for extracting message content from the received messages;

a message intake manager implemented on one of the plurality of processors for collecting recipient data available to the network by searching for social data, spatial data, temporal data and logical data relating to the at least one recipient and the extracted message content;

a correlation engine implemented on one of the plurality of processors for correlating relationships between the extracted message content and the at least one recipient associated with the messages;

a cluster building engine implemented on one of the plurality of processors for forming clusters via grouping the messages based upon the correlated relationships between the extracted message content and the at least one recipient;

a summarization engine implemented on one of the plurality of processors for summarizing the content of the messages contained within each of the clusters formed by the cluster building engine; and

a labeling engine implemented on one of the plurality of processors for analyzing the summarized cluster content produced by the summarization engine to determine a label, wherein the label provides a visual cluster summarization representation for the at least one recipient of the clusters.

32. The system of claim 31 further comprising:

a cluster action engine implemented on one of a plurality of processors for monitoring each cluster for actions taken on or with the cluster label, wherein upon detection of said action, requesting feedback in view of the clusters.

33. The system of claim 31 further comprising:

a content server implemented on a third party processor for producing additional content for input into the message, wherein the additional content is embedded in the message in exchange for payment to the third party.

34. The system of claim 33, wherein the content server interacts directly with the cluster building engine during the forming of the clusters of the related messages.

35. The system of claim 31 further comprising:

a message delivery manager implemented on one of the plurality of processors for determining if a delivery condition is included in the clusters and communicating the clusters to the at least one recipient in accordance with the delivery condition being met.