Processing method and device of the user input information

• US 10,503,729 B2
• Filed: 03/27/2014
• Issued: 12/10/2019
• Est. Priority Date: 06/20/2013
• Status: Active Grant

• Current Assignee: Tencent Technology Company Limited (Tencent Holdings Limited)
• Original Assignee: Tencent Technology Company Limited (Tencent Holdings Limited)
• Inventors: Wang, Fei, Xu, Hao, Zhu, Pan, Han, Guiping, Liao, Ling
• Primary Examiner(s): Harper, Eliyah S.
• Application Number:
  US14/228,149
• Publication Number:
  US 20140379738A1
• Time in Patent Office: 2,084 Days
• Field of Search: 707786, 707736
• CPC Class Codes: G06F 16/24 Querying
• CPC Class Codes: G06F 16/3329 Natural language query formulation or dialogue systems
• CPC Class Codes: G06F 16/3344 using natural language analysis

Abstract

A method and device for processing user input, including: storing a respective usage rule for processing a user request for a task, the respective usage rule comprising a respective rule string, a respective command index number, and a respective parameter string; performing a first transformation on the respective rule string of the respective usage rule to obtain a respective regular expression pattern; performing a second transformation on the respective rule string of the respective usage rule to obtain a respective standard regular expression; obtaining a user input that matches the respective rule string of the respective usage rule; obtaining a concatenated parameter string using the respective regular expression pattern and the respective standard regular expression; and executing a command corresponding to the respective command index number using the concatenated parameter string.

Claims

1. A method of processing Chinese language user input during a natural language man-machine interaction, comprising:

at a device having one or more processors and memory;

storing a respective usage rule for processing a Chinese natural language user request for a computer-executable task, wherein the computer-executable task corresponds to multiple alternative natural language expressions in the Chinese language, wherein the respective usage rule comprises a respective rule string that includes the multiple alternative natural language expressions in Chinese, at least one placeholder for inserting (i) a respective named entity in the respective rule string in accordance with Chinese grammatical rules, (ii) a respective command index number that corresponds to a command to process the computer-executable task, and (iii) a respective parameter string that identifies respective types of input values to be obtained from the user request and used to process the computer-executable task;

performing a first transformation on the respective rule string of the respective usage rule to obtain a respective regular expression pattern by converting one or more Chinese characters in the respective rule string into a first identifier, wherein the first identifier includes a regular expression for a respective string;

performing a second transformation on the respective rule string of the respective usage rule to obtain a respective standard regular expression by adding a second identifier immediately before the at least one placeholder and another instance of the second identifier immediately after the at least one placeholder in the respective rule string, wherein;

the second identifier includes a regular expression for a wildcard;

the respective regular expression pattern obtained from the respective rule string is different from the respective standard regular expression obtained from the respective rule string, and

the respective regular expression pattern obtained through the first transformation is the same as the respective standard regular expression obtained through the second transformation when the respective regular expression pattern obtained through the first transformation is filled with one or more correct named entities recognized from a Chinese language user input, at respective locations in the respective regular expression pattern corresponding to the at least one placeholder specified in the respective rule string;

receiving a speech input from a user using a microphone installed in the device;

converting the speech input into a Chinese text string through speech-to-text processing as a first text input in Chinese;

in accordance with a determination that the first user input in Chinese matches at least one of the multiple alternative natural language expressions in the respective rule string of the respective usage rule;

obtaining a concatenated parameter string using the respective regular expression pattern of the respective usage rule obtained through the first transformation and the respective standard regular expression of the respective usage rule obtained through the second transformation, including;

extracting one or more named entities from the first user input in Chinese; and

inserting the one or more named entities recognized from the first user input at locations of corresponding parameters in the concatenated parameter string of the respective usage rule; and

performing the computer-executable task by executing the command corresponding to the respective command index number using the concatenated parameter string and producing a result from the execution of the command.

2. The method of claim 1, wherein obtaining the concatenated parameter string using the respective regular expression pattern and the respective standard regular expression further comprises:

in accordance with a determination that the first user input matches the respective rule string of the respective usage rule, extracting a respective intermediate parameter from the first user input in accordance with the respective regular expression pattern, the extracting leaving a remainder portion of the first user input;

performing named entity recognition for the respective intermediate parameter to obtain at least one named entity result;

generating at least one intermediate result by concatenating the recognized named entity result with the remainder portion of the first user input;

performing matching analysis between the at least one intermediate result and the respective standard regular expression to determine a matched named entity result; and

obtaining actual parameter values for the concatenated parameter string based on one or more respective named entities identified in accordance with the matched named entity result.

3. The method of claim 2, wherein obtaining the concatenated parameter string using the respective regular expression pattern and the respective standard regular expression further comprises:

constructing the concatenated parameter string in accordance with the respective parameter string of the respective usage rule and the one or more respective named entities identified in accordance with the matched named entity result.

4. The method of claim 2, wherein performing the first transformation on the respective rule string of the respective usage rule to obtain the respective regular expression pattern further comprises:

identifying content between a first parameter and a second parameter in the respective rule string of the respective usage rule; and

generating the respective regular expression pattern for the respective usage rule by substituting the content between the first parameter and the second parameter with the first identifier.

5. The method of claim 4, wherein extracting the respective intermediate parameter from the first user input in accordance with the respective regular expression pattern further comprises:

extracting the respective intermediate parameter from the first user input at a location indicated by the first identifier.

6. The method of claim 2, wherein performing the second transformation on the respective rule string of the respective usage rule to obtain the respective standard regular expression further comprises:

identifying all parameters in the respective rule string of the respective usage rule; and

generating the respective standard regular expression for the respective usage rule by inserting the second identifier before and after each parameter identified in the respective rule string of the respective usage rule.

7. The method of claim 6, wherein performing the matching analysis between the at least one intermediate result and the respective standard regular expression further comprises:

determining whether each of the at least one intermediate result matches the respective standard regular expression; and

in accordance with a determination that a respective one of the at least one intermediate result matches the respective standard regular expression, using the respective intermediate result as the matched intermediate result.

8. The method of claim 1, further comprising:

storing a plurality of usage rules each having a respective priority level; and

processing the first user input against respective regular expression patterns of the plurality of usage rules in an order according to descending priority levels of the plurality of usage rules to determine whether the first user input matches the respective regular expression pattern of the respective usage rule.

9. A system of processing user input, comprising:

one or more processors; and

memory having instructions stored thereon, the instructions, when executed by the one or more processors, cause the processors to perform operations comprising;

storing a respective usage rule for processing a Chinese natural language user request for a computer-executable task, wherein the computer-executable task corresponds to multiple alternative natural language expressions in the Chinese language, wherein the respective usage rule comprises a respective rule string that includes the multiple alternative natural language expressions in Chinese, at least one placeholder for inserting a respective named entity in the respective rule string in accordance with Chinese grammatical rules, a respective command index number that corresponds to a command to process the computer-executable task, and a respective parameter string that identifies respective types of input values to be obtained from the user request and used to process the computer-executable task;

performing a first transformation on the respective rule string of the respective usage rule to obtain a respective regular expression pattern by converting one or more Chinese characters in the respective rule string into a first identifier, wherein the first identifier includes a regular expression for a respective string;

performing a second transformation on the respective rule string of the respective usage rule to obtain a respective standard regular expression by adding a second identifier immediately before the at least one placeholder and another instance of the second identifier immediately after the at least one placeholder in the respective rule string, wherein;

the second identifier includes a regular expression for a wildcard;

the respective regular expression pattern obtained from the respective rule string is different from the respective standard regular expression obtained from the respective rule string, and

the respective regular expression pattern obtained through the first transformation is the same as the respective standard regular expression obtained through the second transformation when the respective regular expression pattern obtained through the first transformation is filled with one or more correct named entities recognized from a Chinese language user input, at respective locations in the respective regular expression pattern corresponding to the at least one placeholder specified in the respective rule string;

receiving a speech input from a user using a microphone installed in the device;

converting the speech input into a Chinese text string through speech-to-text processing as a first text input in Chinese;

in accordance with a determination that the first user input in Chinese matches at least one of the multiple alternative natural language expressions in the respective rule string of the respective usage rule;

obtaining a concatenated parameter string using the respective regular expression pattern of the respective usage rule obtained through the first transformation and the respective standard regular expression of the respective usage rule obtained through the second transformation, including;

extracting one or more named entities from the first user input in Chinese; and

inserting the one or more named entities recognized from the first user input at locations of corresponding parameters in the concatenated parameter string of the respective usage rule; and

performing the computer-executable task by executing the command corresponding to the respective command index number using the concatenated parameter string and producing a result from the execution of the command.

10. The system of claim 9, wherein obtaining the concatenated parameter string using the respective regular expression pattern and the respective standard regular expression further comprises:

in accordance with a determination that the first user input matches the respective rule string of the respective usage rule, extracting a respective intermediate parameter from the first user input in accordance with the respective regular expression pattern, the extracting leaving a remainder portion of the first user input;

performing named entity recognition for the respective intermediate parameter to obtain at least one named entity result;

generating at least one intermediate result by concatenating the recognized named entity result with the remainder portion of the first user input;

performing matching analysis between the at least one intermediate result and the respective standard regular expression to determine a matched named entity result; and

obtaining actual parameter values for the concatenated parameter string based on one or more respective named entities identified in accordance with the matched named entity result.

11. The system of claim 10, wherein obtaining the concatenated parameter string using the respective regular expression pattern and the respective standard regular expression further comprises:

constructing the concatenated parameter string in accordance with the respective parameter string of the respective usage rule and the one or more respective named entities identified in accordance with the matched named entity result.

12. The system of claim 10, wherein performing the first transformation on the respective rule string of the respective usage rule to obtain the respective regular expression pattern further comprises:

identifying content between a first parameter and a second parameter in the respective rule string of the respective usage rule; and

generating the respective regular expression pattern for the respective usage rule by substituting the content between the first parameter and the second parameter with the first identifier.

13. The system of claim 12, wherein extracting the respective intermediate parameter from the first user input in accordance with the respective regular expression pattern further comprises:

extracting the respective intermediate parameter from the first user input at a location indicated by the first identifier.

14. The system of claim 10, wherein performing the second transformation on the respective rule string of the respective usage rule to obtain the respective standard regular expression further comprises:

identifying all parameters in the respective rule string of the respective usage rule; and

generating the respective standard regular expression for the respective usage rule by inserting the second identifier before and after each parameter identified in the respective rule string of the respective usage rule.

15. The system of claim 14, wherein performing the matching analysis between the at least one intermediate result and the respective standard regular expression further comprises:

determining whether each of the at least one intermediate result matches the respective standard regular expression; and

in accordance with a determination that a respective one of the at least one intermediate result matches the respective standard regular expression, using the respective intermediate result as the matched intermediate result.

16. The system of claim 9, wherein the operations further comprise:

storing a plurality of usage rules each having a respective priority level; and

processing the first user input against respective regular expression patterns of the plurality of usage rules in an order according to descending priority levels of the plurality of usage rules to determine whether the first user input matches the respective regular expression pattern of the respective usage rule.

17. A non-transitory computer-readable medium having instructions stored thereon, the instructions, when executed by one or more processors, cause the processors to perform operations comprising:

storing a respective usage rule for processing a Chinese natural language user request for a computer-executable task, wherein the computer-executable task corresponds to multiple alternative natural language expressions in the Chinese language, wherein the respective usage rule comprises a respective rule string that includes the multiple alternative natural language expressions in Chinese, at least one placeholder for inserting (i) a respective named entity in the respective rule string in accordance with Chinese grammatical rules, (ii) a respective command index number that corresponds to a command to process the computer-executable task, and (iii) a respective parameter string that identifies respective types of input values to be obtained from the user request and used to process the computer-executable task;

performing a first transformation on the respective rule string of the respective usage rule to obtain a respective regular expression pattern, by converting one or more Chinese characters in the respective rule string into a first identifier, wherein the first identifier includes a regular expression for a respective string;

performing a second transformation on the respective rule string of the respective usage rule to obtain a respective standard regular expression by adding a second identifier immediately before the at least one placeholder and another instance of the second identifier immediately after the at least one placeholder in the respective rule string, wherein;

the second identifier includes a regular expression for a wildcard;

the respective regular expression pattern obtained from the respective rule string is different from the respective standard regular expression obtained from the respective rule string, and

the respective regular expression pattern obtained through the first transformation is the same as the respective standard regular expression obtained through the second transformation when the respective regular expression pattern obtained through the first transformation is filled with one or more correct named entities recognized from a Chinese language user input, at respective locations in the respective regular expression pattern corresponding to the at least one placeholder specified in the respective rule string;

receiving a speech input from a user using a microphone installed in the device;

converting the speech input into a Chinese text string through speech-to-text processing as a first text input in Chinese;

in accordance with a determination that the first user input in Chinese matches at least one of the multiple alternative natural language expressions in the respective rule string of the respective usage rule;

obtaining a concatenated parameter string using the respective regular expression pattern of the respective usage rule obtained through the first transformation and the respective standard regular expression of the respective usage rule obtained through the second transformation, including;

extracting one or more named entities from the first user input in Chinese; and

inserting the one or more named entities recognized from the first user input at locations of corresponding parameters in the concatenated parameter string of the respective usage rule; and

performing the computer-executable task by executing the command corresponding to the respective command index number using the concatenated parameter string and producing a result from the execution of the command.

18. The computer-readable medium of claim 17, wherein obtaining the concatenated parameter string using the respective regular expression pattern and the respective standard regular expression further comprises:

in accordance with a determination that the first user input matches the respective rule string of the respective usage rule, extracting a respective intermediate parameter from the first user input in accordance with the respective regular expression pattern, the extracting leaving a remainder portion of the first user input;

performing named entity recognition for the respective intermediate parameter to obtain at least one named entity result;

generating at least one intermediate result by concatenating the recognized named entity result with the remainder portion of the first user input;

performing matching analysis between the at least one intermediate result and the respective standard regular expression to determine a matched named entity result; and

obtaining actual parameter values for the concatenated parameter string based on one or more respective named entities identified in accordance with the matched named entity result.