System and method for semantic processing of natural language commands

Patent metadata

Field	Value
Publication number	US-2016151918-A1
Application number	US-201414557005-A
Country	US
Kind code	A1
Filing date	Dec 1, 2014
Priority date	Dec 1, 2014
Publication date	Jun 2, 2016
Grant date	—

Abstract

A system, method and computer-readable storage devices are for processing natural language commands, such as commands to a robotic arm, using a Tag & Parse approach to semantic parsing. The system first assigns semantic tags to each word in a sentence and then parses the tag sequence into a semantic tree. The system can use statistical approach for tagging, parsing, and reference resolution. Each stage can produce multiple hypotheses, which are re-ranked using spatial validation. Then the system selects a most likely hypothesis after spatial validation, and generates or outputs a command. In the case of a robotic arm, the command is output in Robot Control Language (RCL).

First claim

We claim:

1 . A method comprising:
receiving a natural language command; assigning, in a first stage, semantic tags to the natural language command to yield a tagged natural language command; semantically parsing, in a second stage, the tagged natural language command to yield a parsed natural language command; and identifying a command type for the parsed natural language command.

2 . The method of claim 1 , wherein the semantic tags identify entity types and event types in the input command.

3 . The method of claim 1 , further comprising, after identifying the command type:
performing spatial validation based on the command type for a context of the natural language command; and if the spatial validation indicates that certainty of the command type identified meets a certainty threshold, executing the natural language command; and if the spatial validation indicates that certainty of the command type identified does not meet the certainty threshold, prompting for clarification of the natural language command.

4 . The method of claim 3 , wherein the spatial validation is based on a condition of a target area associated with the natural language command.

5 . The method of claim 4 , wherein the target area comprises a working area of a robotic arm, and wherein the condition of the target area comprises presence and positions of objects in the working area.

6 . The method of claim 5 , further comprising:
performing reference resolution on the natural language command to uniquely identify one of the objects in the working area.

7 . The method of claim 1 , wherein the natural language command is directed to one of a physical robotic appendage or a virtual robotic appendage.

8 . The method of claim 7 , further comprising outputting the parsed natural language command according to the command type and according to a robot control language.

9 . The method of claim 1 , wherein the semantic tags are assigned using a maximum entropy sequence tagger.

10 . The method of claim 1 , wherein a semantic parser trained on a non-lexical semantic tree parses the tagged natural language command.

11 . A system comprising:
a processor; and a computer-readable storage medium having instructions stored which, when executed by the processor, cause the processor to perform operations comprising:
receiving a natural language command;
assigning, in a first stage, semantic tags to the natural language command to yield a tagged natural language command;
semantically parsing, via a parsing processor and in a second stage, the tagged natural language command to yield a parsed natural language command; and
identifying a command type for the parsed natural language command.

12 . The system of claim 11 , wherein the semantic tags identify entity types and event types in the input command.

13 . The system of claim 11 , the computer-readable storage medium having additional instructions stored which, when executed by the processor, cause the processor to perform operations comprising, after identifying the command type:
performing spatial validation based on the command type for a context of the natural language command; and if the spatial validation indicates that certainty of the command type identified meets a certainty threshold, executing the natural language command; and if the spatial validation indicates that certainty of the command type identified does not meet the certainty threshold, prompting for clarification of the natural language command.

14 . The system of claim 13 , wherein the spatial validation is based on a condition of a target area associated with the natural language command.

15 . The system of claim 14 , wherein the target area comprises a working area of a robotic arm, and wherein the condition of the target area comprises presence and positions of objects in the working area.

16 . The system of claim 15 , the computer-readable storage medium having additional instructions stored which, when executed by the processor, cause the processor to perform operations comprising:
performing reference resolution on the natural language command to uniquely identify one of the objects in the working area.

17 . A computer-readable storage device having instructions stored which, when executed by a computing device, cause the computing device to perform operations comprising:
receiving a natural language command; assigning, in a first stage, semantic tags to the natural language command to yield a tagged natural language command; semantically parsing, in a second stage, the tagged natural language command to yield a parsed natural language command; and identifying a command type for the parsed natural language command.

18 . The computer-readable storage device of claim 17 , wherein the natural language command is directed to one of a physical robotic appendage or a virtual robotic appendage.

19 . The computer-readable storage device of claim 18 , having additional instructions stored which, when executed by the computing device, cause the computing device to perform operations comprising:
outputting the parsed natural language command according to the command type and according to a robot control language.

20 . The computer-readable storage device of claim 17 , wherein the semantic tags are assigned using a maximum entropy sequence tagger.

Assignees

• At & T Ip I Lp

Inventors

• Stoyanchev Svetlana
• Bangalore Srinivas
• Chen John
• Jung Hyuckchul

Classifications

• B25J13/003 Primary

  by means of an audio-responsive input (audible safety signals B25J19/061) · CPC title

• G06F40/30

  Semantic analysis · CPC title

• G06F17/28

  Physics · mapped topic

• G06F17/2785

  Physics · mapped topic