Data mining electronic communications

• US 9,529,864 B2
• Filed: 08/28/2009
• Issued: 12/27/2016
• Est. Priority Date: 08/28/2009
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Inventors: Edelen, James J., Sundelin, Tore L., Kleewein, James C., Clark, Bradford R., Pereira, Jorge
• Primary Examiner(s): Ashraf, Waseem
• Application Number:
  US12/550,189
• Publication Number:
  US 20110055196A1
• Time in Patent Office: 2,678 Days
• Field of Search: 709206-207
• US Class Current: 1/1
• CPC Class Codes: G06F 16/24575 using context
• CPC Class Codes: G06F 16/2465 Query processing support for facilitating data mining operations in structured databases

Abstract

User insights derived from communication data may be provided. Communications associated with a user may be collected and analyzed to derive insights about the user. The insight may then be provided to the user, such as for modifying an application functionality, creating a message processing rule, providing new information to the user, and updating an application display or user interface.

Claims

1. A method for providing electronic communication data mining, the method comprising:

collecting a plurality of communications associated with a user, wherein the plurality of communications comprise a plurality of data types comprising at least two of the following;

an e-mail, an instant messages (IM), a short message service (SMS) message, a voicemail, a search query, a public posting, a document authored by the user, a directory of a plurality of users, a contact list, a calendar appointment, and a task entry;

delivering the plurality of collected communications to at least one data analyzer module, the at least one data analyzer module comprising a plurality of weighted workflows, each of the plurality of weighted workflows comprising a different level of processing for generating insights and updating insights based on at least one of batch data, collected data and generated data, the plurality of weighted workflows comprising at least a medium weight workflow for updating existing insights based on the at least one of the batch data and a lightweight workflow for at least one of generating and updating the insights in real-time, the plurality of workflows being executed on a plurality of machines in a high availability cluster, the plurality of machines including at least one active node utilized by the lightweight workflow and one passive node utilized by the medium weight workflow, the medium weight workflow utilizing a resource intensive batch analysis that is executed using spare cycles of the passive node;

deriving, utilizing one or more of the plurality of weighted workflows, an insight about the user from the collected plurality of communications, the insight about the user being based on at least one of the following;

a processor type used to derive the insight and a communication type for at least one of the collected communications from which the insight is derived; and

modifying at least one application functionality based on the derived insight.

2. The method of claim 1, wherein the derived insight comprises at least one of the following:

a habit, a preference, an interest, a relationship, a prioritization, an area of expertise, and a demographic factor, an active topic, and a triage behavior.

3. The method of claim 1, wherein the derived insight is stored on a server accessible by the user from a plurality of clients.

4. The method of claim 1, further comprising assigning a confidence to the derived insight.

5. The method of claim 1, further comprising updating the derived insight according to analysis of at least one of the following:

a newly received communication and a newly sent communication.

6. The method of claim 1, further comprising receiving an edit to the derived insight.

7. The method of claim 6, wherein the edit to the derived insight is received from at least one of the following:

the user and an administrator.

8. The method of claim 1, wherein modifying the at least one application comprises at least one of the following:

creating a contact group, re-ordering a message list, applying a visual property to a message associated with the insight, creating a processing rule to be applied to incoming communications, creating an alert associated with a communication associated with another user, creating a processing rule for future communications, modifying a processing rule for future communications, reorganizing stored data associated with the application, and creating an association of a plurality of data associated with the application.

9. The method of claim 1, further comprising displaying an explanation of the derived insight to the user.

10. The method of claim 9, wherein displaying the explanation of the derived insight to the user comprises displaying at least one factor used to derive the insight and a user interface element operative to edit the at least one factor used to derive the insight.

11. The method of claim 1, further comprising cleansing the collected plurality of communications prior to analysis.

12. The method of claim 1, wherein deriving the insight comprises extracting, from at least one of the plurality of communications, at least one of the following:

a name, a phone number, a key word, a contact address, an urgency, an associated topic, a subject area, a type of communication, a date, a user action, a user feedback, a relationship between the user and a sender of the at least one of the plurality of communications.

13. A tangible computer-readable device which stores a set of instructions which when executed performs a method for providing communication data mining, the method executed by the set of instructions comprising:

analyzing a plurality of communications associated with a user utilizing at least one data analyzer module, the at least one data analyzer module comprising a plurality of weighted workflows, each of the plurality of weighted workflows comprising a different level of processing for generating insights and updating insights based on at least one of batch data, collected data and generated data, the plurality of weighted workflows comprising at least a medium weight workflow for updating existing insights based on the at least one of the batch data and a lightweight workflow for at least one of generating and updating the insights in real-time, the plurality of workflows being executed on a plurality of machines in a high availability cluster, the plurality of machines including at least one active node utilized by the lightweight workflow and one passive node utilized by the medium weight workflow, the medium weight workflow utilizing a resource intensive batch analysis that is executed using spare cycles of the passive node, wherein the plurality of communications comprise a plurality of data types comprising at least two of the following;

an e-mail, an instant messages (IM), a short message service (SMS) message, a voicemail, a search query, a public posting, a document authored by the user, a directory of a plurality of users, a contact list, a calendar appointment, and a task entry;

deriving an insight based on the analyzed plurality of communications; and

providing the insight to the user when the insight based on at least one of the following;

a processor type used to derive the insight and a communication type for at least one of the collected communications from which the insight is derived, wherein providing the insight to the user comprises at least one of the following;

creating a message processing rule, providing at least one new piece of information to the user, updating an application display, and adding at least one functionality to an application.

14. The tangible computer-readable device of claim 13, further comprising storing the insight on a server accessible by the user from a plurality of locations.

15. The tangible computer-readable device of claim 13, wherein creating a message processing rule comprises:

deriving the insight using a property associated with the plurality of communications,

creating the message processing rule according to the derived insight to be applied to at least one newly received communication, and

modifying the at least one newly received communication, wherein the at least one newly received communication matches the property used to derive the insight.

16. The tangible computer-readable device of claim 15, wherein the property used to derive the insight comprises at least one of the following:

a sender, a recipient, a subject, a time, a priority, and a keyword.

17. The tangible computer-readable device of claim 15, wherein providing the insight to the user comprises sorting the plurality of communications for display according to a criteria of the derived insight.

18. The tangible computer-readable device of claim 15, wherein providing the insight to the user comprises displaying a longer preview for a preview of a first communication of the plurality of communications and a shorter preview of a second communication of the plurality of communications.

19. A system for providing user insights through data mining, the system comprising:

a memory storage; and

a processing unit coupled to the memory storage, wherein the processing unit is operative to;

collect a plurality of data elements from a plurality of sources associated with a user, wherein the plurality of sources comprises at least two of the following;

a document, an e-mail, an SMS message, a voicemail, a search query, a web submission, a forum posting, a blog posting, a calendar item, a to-do list entry, an instant message, and a directory;

cleanse at least one data element of the plurality of data elements, wherein being operative to cleanse the at least one data element comprises at least one of the following;

being operative to perform a spell check on the at least one data element, validate the at least one data element, normalize the at least one data element, correct the at least one data element, and remove at least one extraneous character from the at least one data element;

extract at least one piece of key information from the at least one data element, wherein the at least one piece of key information comprises at least one of the following;

a sender, a recipient, a time, a subject, a contact method, a related communication, and a storage location;

analyze the at least one piece of key information utilizing at least one data analyzer module comprising a plurality of weighted workflows, each of the plurality of weighted workflows comprising a different level of processing for generating insights and updating insights based on at least one of batch data, collected data and generated data, the plurality of weighted workflows comprising at least a medium weight workflow for updating existing insights based on the at least one of the batch data and a lightweight workflow for at least one of generating and updating the insights in real-time, the plurality of workflows being executed on a plurality of machines in a high availability cluster, the plurality of machines including at least one active node utilized by the lightweight workflow and one passive node utilized by the medium weight workflow, the medium weight workflow utilizing a resource intensive batch analysis that is executed using spare cycles of the passive node, the analyzed at least one piece of key information being utilized to derive an insight associated with the user, wherein the derived insight comprises at least one of the following;

a behavior, a pattern, a relationship, an interest, a preference, a demographic factor, and an expertise;

assign a confidence probability to the insight;

determine whether the assigned confidence probability of the insight meets a threshold level to provide the insight to the user, the threshold level to provide the insight being based on at least one of the following;

a processor type used to derive the insight and a communication type for at least one of the collected communications from which the insight is derived;

in response to determining that the assigned confidence probability of the insight is high enough to provide the insight to the user, provide the insight to the user wherein being operative to provide the insight to the user comprises at least one of the following;

creating a message processing rule, providing at least one new piece of information to the user, updating an application display, and adding at least one functionality to an application;

store the insight in a location accessible by the user from a plurality of clients; and

receive at least one modification to the insight from the user, wherein the at least one modification comprises at least one of the following;

a feedback rating, a disabling of a use of at least one of the plurality of data elements, an enabling of a use of at least one of the plurality of data elements, a disabling of the insight, an enabling of the insight, a weighting of at least one of the plurality of data elements, and a modification of the confidence probability.