System, method, and recording medium for corpus curation for action manifestation for cognitive robots

• US 10,239,205 B2
• Filed: 06/29/2016
• Issued: 03/26/2019
• Est. Priority Date: 06/29/2016
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Rusu, Laura Irina, Sharon, Guy, Sivakumar, Gandhi
• Primary Examiner(s): Bendidi, Rachid
• Application Number:
  US15/196,148
• Publication Number:
  US 20180001470A1
• Time in Patent Office: 1,000 Days
• CPC Class Codes: B25J 11/0005 Manipulators having means for high-level communication with users, e.g. speech generator, face recognition means
• CPC Class Codes: B25J 9/163 learning, adaptive, model based, rule based expert control
• CPC Class Codes: G06N 20/00 Machine learning
• CPC Class Codes: G06N 3/008 based on physical entities controlled by simulated intelligence so as to replicate intelligent life forms, e.g. based on robots replicating pets or humans in their appearance or behaviour
• CPC Class Codes: G06N 5/022 Knowledge engineering; Knowledge acquisition
• CPC Class Codes: G06N 5/046 Forward inferencing; Production systems
• CPC Class Codes: Y10S 901/03 Teaching system

Abstract

A corpus curation method, system, and non-transitory computer readable medium, include mapping a kinematic motion of a robot to a granular feature of an item in the corpus and answering a user question using the mapped kinematic motion embedded in an answer by the robot.

Claims

1. A corpus curation method including a corpus, the method comprising:

mapping, via a processor, a kinematic motion of a robot to a granular feature of an item in the corpus to create an offline database of the kinematic motion mapped to the granular feature of the item or of a kinematic motion mapped to a combination of two granular features of the item; and

answering, via the processor, a user question using the mapped kinematic motion in the offline database embedded in an answer by the robot,

wherein the granular feature of the item in the corpus comprises a feature at a lowest granular level at which the feature cannot be broken down further,

wherein a delivery type of the answer by the robot includes at least one of;

a textual display interface for a text output;

a kinematic motion output by the robot; and

a speech output by the robot, and

wherein an audience type comprises any of;

a blind user;

a hearing-impaired user; and

a non-impaired user; and

the method further comprising determining the audience type and the delivery capability and causing the robot to change the answer based on the delivery capability and the audience type,

wherein, if the determining determines the audience type as the hearing-impaired user, the answering only causes the robot to deliver the answer with the kinematic motion, and

wherein, if the determining determines the audience type as the blind user, the answering only causes the robot to deliver the answer with a speech output.

2. The method of claim 1, wherein the granular feature comprises a value of a feature of the item in the corpus.

3. The method of claim 1, wherein the mapping maps a plurality of kinematic motions to the granular feature.

4. The method of claim 1, wherein the kinematic motion comprises a visual representation of a value of the granular feature such that the robot performs the kinematic motion during the answering as the visual representation of the item in the answer.

5. The method of claim 1, further comprising determining an audience type of the answer and a delivery capability of the robot and causing the answering to change a context of the answer according to the audience and the delivery capability.

6. The method of claim 1, further comprising inferring the kinematic motion of a mapped item to a related item in the corpus when the related item is not mapped by the mapping such that answering causes the robot to perform the kinematic motion of the mapped item for the related item.

7. The method of claim 6, wherein the inferring determines when the related item is related to the granular feature of a mapped item by an inbuilt ontology in the corpus.

8. The method of claim 6, wherein the inferring uses an association rule by data mining the corpus to infer the related item to a mapped item.

9. The method of claim 1, further comprising inferring a velocity of the kinematic motion performed by the robot based on a velocity feature indicating a speed of the granular feature of the item in the corpus.

10. The method of claim 1, wherein, if the audience type is determined as the non-impaired user, the answering answers with any one of the textual display interface for the text output, the kinematic motion output by the robot, and the speech output by the robot.

11. The method of claim 1, wherein, if the audience type is determined as the non-impaired user, the answering answers with a combination of the textual display interface for the text output, the kinematic motion output by the robot, and the speech output by the robot.

12. The method of claim 1, wherein, if the audience type is determined as the non-impaired user, the answering answers with all of the textual display interface for the text output, the kinematic motion output by the robot, and the speech output by the robot.

13. The method of claim 1, wherein, if the audience type is determined as the non-impaired user, the answering answers with at least two of the textual display interface for the text output, the kinematic motion output by the robot, and the speech output by the robot.

14. A non-transitory computer-readable recording medium recording a corpus curation program including a corpus, the program causing a computer to perform:

mapping a kinematic motion of a robot to a granular feature of an item in the corpus to create an offline database of the kinematic motion mapped to the granular feature of the item or of a kinematic motion mapped to a combination of two granular features of the item; and

answering a user question using the mapped kinematic motion in the offline database embedded in an answer by the robot,

wherein the granular feature of the item in the corpus comprises a feature at a lowest granular level at which the feature cannot be broken down further,

wherein a delivery type of the answer by the robot includes at least one of;

a textual display interface for a text output;

a kinematic motion output by the robot; and

a speech output by the robot, and

wherein an audience type comprises any of;

a blind user;

a hearing-impaired user; and

a non-impaired user, and

further comprising determining the audience type and the deliver capability and causing the robot to change the answer based on the delivery capability and the audience type,

wherein, if the determining determines the audience type as the hearing-impaired user, the answering only causes the robot to deliver the answer with the kinematic motion, and

wherein if the determining determines the audience type as the blind user, the answering only causes the robot to deliver the answer with a speech output.

15. The non-transitory computer-readable recording medium of claim 14, wherein the granular feature comprises a value of a feature of the item in the corpus.

16. The non-transitory computer-readable recording medium of claim 14, wherein the mapping maps a plurality of kinematic motions to the granular feature.

17. The non-transitory computer-readable recording medium of claim 14, wherein the kinematic motion comprises a visual representation of a value of the granular feature such that the robot performs the kinematic motion during the answering as the visual representation of the item in the answer.

18. The non-transitory computer-readable recording medium of claim 14, further comprising determining an audience type of the answer and a delivery capability of the robot and causing the answering to change a context of the answer according to the audience and the delivery capability.

19. The non-transitory computer-readable recording medium of claim 14, further comprising inferring the kinematic motion of a mapped item to a related item in the corpus when the related item is not mapped by the mapping such that the answering causes the robot to perform the kinematic motion of the mapped item for the related item.

20. A corpus curation system, said system comprising:

a corpus;

a processor; and

a memory, the memory storing instructions to cause the processor to;

map a kinematic motion of a robot to a granular feature of an item in the corpus to create an offline database of the kinematic motion mapped to the granular feature of the item or of a kinematic motion mapped to a combination of two granular features of the item; and

answer a user question using the mapped kinematic motion in the offline database embedded in an answer by the robot,

wherein the granular feature of the item in the corpus comprises a feature at a lowest granular level at which the feature cannot be broken down further,

wherein a delivery type of the answer by the robot includes at least one of;

a textual display interface for a text output;

a kinematic motion output by the robot; and

a speech output by the robot, and

wherein an audience type comprises any of;

a blind user;

a hearing-impaired user; and

a non-impaired user, and

the memory further storing instructions to cause the processor to determine the audience type and the delivery capability and causing the robot to chance the answer based on the delivery capability and the audience type,

wherein, if the audience type is determined as the hearing-impaired user, the answering only causes the robot to deliver the answer with the kinematic motion, and

wherein, if the audience type is determined as the blind user, the answering only causes the robot to deliver the answer with a speech output.