Systems and methods for providing input to virtual agent

• US 9,560,089 B2
• Filed: 12/11/2012
• Issued: 01/31/2017
• Est. Priority Date: 12/11/2012
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Nuance Communications, Inc. (Microsoft Corporation)
• Inventors: Lynch, Timothy, Harper, Kenneth S., Radebaugh, Carey
• Primary Examiner(s): Hussain, Tauqir
• Assistant Examiner(s): MOREAU, AUSTIN J
• Application Number:
  US13/710,721
• Publication Number:
  US 20140164509A1
• Time in Patent Office: 1,512 Days
• US Class Current: 1/1
• CPC Class Codes: H04L 65/403 Arrangements for multi-party communication, e.g. for conferences data switching systems for conference H04L12/18; arrangements for connecting several subscribers to a common circuit, i.e. affording conference facilities H04M3/56; television conferencing systems H04N7/15
• CPC Class Codes: H04L 67/30 Profiles
• CPC Class Codes: H04L 67/52 specially adapted for the location of the user terminal

Abstract

Systems, methods, and apparatus for use in connection with at least one virtual agent. In some embodiments, at least one processor is programmed to use a record of a multiparty conversation to provide input to the at least one virtual agent.

Claims

1. An apparatus comprising:

at least one memory storing one or more executable instructions; and

at least one hardware processor coupled to the at least one memory and programmed by the one or more executable instructions to implement at least one virtual agent, the at least one virtual agent being programmed to;

divide a record of a multiparty conversation into a plurality of segments, wherein the multiparty conversation comprises a prior virtual agent interaction during which at least one first task was performed by the at least one virtual agent;

assign weights to at least some of the plurality of segments based on a chronological ordering of the plurality of segments, wherein the weights are determined by statistical analysis of a corpus of conversation histories tagged with respective user intents, a first weight assigned to a first segment is lower than a second weight assigned to a second segment that is later in time than the first segment, and the plurality of segments have successively increasing lengths from segment that is latest in time to a segment that is earliest in time; and

analyze the at least some of the plurality of segments to identify at least one second task to be performed, wherein the at least one virtual agent is programmed to prioritize information obtained from the at least some of the plurality of segments based on the respective weights.

2. The apparatus of claim 1, wherein the length of each segment is based on an amount of text in the segment.

3. The apparatus of claim 1, wherein the at least one processor is programmed to modify the record of the multiparty conversation and use the modified record to provide input to the at least one virtual agent.

4. The apparatus of claim 3, wherein modifying the record of the multiparty conversation comprises removing at least one portion of the multiparty conversation.

5. The apparatus of claim 1, wherein the at least one virtual agent is programmed to:

identify user preference information based on the record of the multiparty conversation; and

make a recommendation by taking into account the user preference information.

6. The apparatus of claim 1, wherein the multiparty conversation is between at least two persons, and wherein the at least one second task to be performed comprises making a joint recommendation for the at least two persons, the joint recommendation relating to a social gathering to be attended by the at least two persons.

7. The apparatus of claim 1, wherein the plurality of segments correspond, respectively, to different windows of time, and wherein the length of each segment is based on a length of the corresponding window of time.

8. A method performed by at least one hardware processor in connection with at least one virtual agent, the method comprising acts of:

dividing a record of a multiparty conversation into a plurality of segments, wherein the multiparty conversation comprises a prior virtual agent interaction during which at least one first task was performed by the at least one virtual agent;

assigning weights to at least some of the plurality of segments based on a chronological ordering of the plurality of segments, wherein the weights are determined by statistical analysis of a corpus of conversation histories tagged with respective user intents, a first weight assigned to a first segment is lower than a second weight assigned to a second segment that is later in time than the first segment, and the plurality of segments have successively increasing lengths from a segment that is latest in time to a segment that is earliest in time; and

analyzing the at least some of the plurality of segments to identify at least one second task to be performed, wherein the at least one virtual agent is programmed to prioritize information obtained from the at least some of the plurality of segments based on the respective weights.

9. The method of claim 8, wherein the length of each segment is based on an amount of text in the segment.

10. The method of claim 8, further comprising modifying the record of the multiparty conversation, wherein the modified record is used to provide input to the at least one virtual agent.

11. The method of claim 10, wherein modifying the record of the multiparty conversation comprises removing at least one portion of the multiparty conversation.

12. The method of claim 8, further comprising:

identifying user preference information based on the record of the multiparty conversation; and

making a recommendation by taking into account the user preference information.

13. The method of claim 8, wherein the multiparty conversation is between at least two persons, and wherein the at least one second task to be performed comprises making a joint recommendation for the at least two persons, the joint recommendation relating to a social gathering to be attended by the at least two persons.

14. The method of claim 8, wherein the plurality of segments correspond, respectively, to different windows of time, and wherein the length of each segment is based on a length of the corresponding window of time.

15. At least one computer-readable storage device having encoded thereon instructions that, when executed by at least one processor, perform a method for use in connection with at least one virtual agent, the method comprising acts of:

dividing a record of a multiparty conversation into a plurality of segments, wherein the multiparty conversation comprises a prior virtual agent interaction during which at least one first task was performed by the at least one virtual agent;

assigning weights to at least some of the plurality of segments based on a chronological ordering of the plurality of segments, wherein the weights are determined by statistical analysis of a corpus of conversation histories tagged with respective user intents, a first weight assigned to a first segment is lower than a second weight assigned to a second segment that is later in time than the first segment, and the plurality of segments have successively increasing lengths from a segment that is latest in time to a segment that is earliest in time; and

analyzing the at least some of the plurality of segments to identify at least one second task to be performed, wherein the at least one virtual agent is programmed to prioritize information obtained from the at least some of the plurality of segments based on the respective weights.

16. The at least one computer-readable storage device of claim 15, wherein the length of each segment is based on an amount of text in the segment.

17. The at least one computer-readable storage device of claim 15, wherein the method further comprises modifying the record of the multiparty conversation, wherein the modified record is used to provide input to the at least one virtual agent.

18. The at least one computer-readable storage device of claim 17, wherein modifying the record of the multiparty conversation comprises removing at least one portion of the multiparty conversation.

19. The at least one computer-readable storage device of claim 15, wherein the method further comprises:

identifying user preference information based on the record of the multiparty conversation; and

making a recommendation by taking into account the user preference information.

20. The at least one computer-readable storage device of claim 15, wherein the multiparty conversation is between at least two persons, and wherein the at least one second task to be performed comprises making a joint recommendation for the at least two persons, the joint recommendation relating to a social gathering to be attended by the at least two persons.

21. The at least one computer-readable storage device of claim 15, wherein the plurality of segments correspond, respectively, to different windows of time, and wherein the length of each segment is based on a length of the corresponding window of time.