WO2017084185A1

WO2017084185A1 - Intelligent terminal control method and system based on semantic analysis, and intelligent terminal

Info

Publication number: WO2017084185A1
Application number: PCT/CN2015/099908
Authority: WO
Inventors: 管继炜; 林天麟; 肖杰; 唐志海; 斯美樑; 余刚; 杨帆
Original assignee: 深圳创想未来机器人有限公司
Priority date: 2015-11-19
Filing date: 2015-12-30
Publication date: 2017-05-26
Also published as: CN105446146A; CN105446146B

Abstract

An intelligent terminal control method based on semantic analysis, comprising the following steps: providing a semantic knowledge database, wherein the semantic knowledge database comprises a logical semantic set, the logical semantic set comprises a logical semantic word, the logical semantic word comprises a condition word and an execution word; receiving an input signal, wherein the input signal is a voice signal or a picture signal; identifying a character through the voice signal or the picture signal and storing the identified character into a target text; identifying the condition word and the execution word in the target text; based on the condition word and the execution word, analysing a perception condition and an execution command; and executing the execution command according to the perception condition. The embodiments of the present invention further relate to an intelligent terminal control system based on semantic analysis and an intelligent terminal.

Description

Intelligent terminal control method, system and intelligent terminal based on semantic analysis

[0001] The present invention relates to the field of intelligent terminal control, and in particular to an intelligent terminal control method, system and intelligent terminal based on semantic analysis.

Background technique

[0002] In recent years, due to breakthroughs in the field of speech recognition technology, speech interaction has moved from the laboratory to the application. The human-machine mode of voice interaction has been applied in some scenarios. Natural language understanding in speech recognition technology is a technique for translating and allowing a machine to understand a user's intention, identify a user's language content, and perform a specific operation in a specific scene. As a technology, semantic understanding has a very broad application prospect in smart terminals or services such as smart home appliances, service robots, mobile terminals, teleconferencing, smart customer service and wearable devices.

[0003] In today's Internet information development, the home Internet of Things is entering the ordinary family. As the mainstream control method, voice intelligent control technology has been popularized in family life. However, existing voice solutions and technologies lack matching application scenarios for family life, which makes semantic understanding very limited, and existing semantic understanding often cannot solve logical judgment problems.

technical problem

In view of the above, it is necessary to provide an intelligent terminal control method, system and intelligent terminal based on semantic analysis and having logical judgment capabilities.

Problem solution

Technical solution

[0005] A method for controlling an intelligent terminal based on semantic analysis, the method comprising the following steps:

[0006] Providing a semantic knowledge base, the semantic knowledge base comprising a logical semantic set, the logical semantic set comprising a logical semantic word, the logical semantic word comprising a condition word and an execution word;

Receiving an input signal, the input signal being a voice signal or an image signal;

[0008] identifying the text by the voice signal or the image signal, and storing the recognized text in a target text; [0009] identifying the condition word and the execution word in the target text; [0010] analyzing the sensing condition and executing the command according to the condition word and the execution word; and

[0011] executing the execution command according to the sensing condition.

[0012] An intelligent terminal control system based on semantic analysis, the system comprising:

[0013] a semantic knowledge base, including a logical semantic set, the logical semantic set includes a logical semantic word, the logical semantic word including a condition word and an execution word;

[0014] a receiving module, configured to receive an input signal, where the input signal is a voice signal or an image signal;

[0015] a voice/image recognition module, configured to identify a text by the voice signal or the image signal, and store the recognized text in a target text;

[0016] an analysis module, configured to identify a condition word and an execution word in the target text, and analyze the sensing condition and the execution command according to the condition word and the execution word;

[0017] an execution module, configured to execute the execution command according to the sensing condition.

[0018] An intelligent terminal includes the above-described intelligent terminal control system based on semantic analysis.

Advantageous effects of the invention

Beneficial effect

[0019] Compared with the prior art, the above-described intelligent terminal control method, system and intelligent terminal based on semantic analysis can perform logical judgment through logical semantic words and quickly obtain structured semantic understanding.

Brief description of the drawing

DRAWINGS

1 is a flowchart of a method for controlling an intelligent terminal based on semantic analysis according to an embodiment of the present invention.

2 is a functional block diagram of a smart terminal control system based on semantic analysis according to an embodiment of the present invention.

3 is a smart terminal control method based on semantic analysis according to another embodiment of the present invention.

Embodiments of the invention

[0023] Please refer to FIG. 1, a method for controlling an intelligent terminal based on semantic analysis in an embodiment of the present invention, including the following steps:

[0024] Step 101: Establish a semantic knowledge base, including a logical semantic set and a scene semantic set. Number of execution objects According to the library, a number of execution objects and a set of functions that can be applied to each execution object are stored.

[0025] The logical semantic set includes logical semantic words, and the logical semantic words include condition words and execution words. Conditional words may include, for example, "if", "if", "if", "as long as" "when...". Execution words may include, for example, "just", "all", etc. for prompting that there will be specific execution command words. The above logical semantic words may also include corresponding other languages, for example, "if...then".

[0026] The scene semantic set includes a perceptual function semantic set and an execution function semantic set. The Perceptual Function Semantic Set stores a number of conditional words, such as time, place, person, sound or specific sound, weather information, and other keywords. Sounds or specific sounds may include onomatopes and animal sounds, and weather information may include temperature, humidity, air pressure, and illuminance. The executive function semantic set includes a number of execution command words, for example, playing audio/video messages, artificial voice reminders, playing music or video, Internet of Things, home appliance control, Internet audio and video calls, basic life questions and answers (such as weather, daytime) Etc.), arrive at a place to find someone and accompany you.

[0027] Executing the object database is to perform a corresponding operation or action, and each execution object is associated with at least one execution command word of the execution function semantic set. Executing the object database can include, for example, managing audio and video players, dynamically triggering smart home controls, life reminder operations, or managing third party applications.

[0028] Step 103: Receive an input signal, where the input signal includes a voice signal or an image signal input by a user. Specifically, the voice signal input by the user may be input by the user through a microphone; the image signal described above may be acquired by a camera.

[0029] Step 105: Identify the input signal, that is, recognize the text by the voice signal or the image signal and store it in a target text, and store the target text and the original voice or image data of the recording user into the target file.

[0030] Step 107: Traverse the target text, identify the keyword to obtain the corresponding semantic information. Each semantic information includes a conditional word and an executive command word. When the semantics are determined, the semantic knowledge base can be quickly searched, and the conditional words and execution command words are extracted according to certain extraction rules, so that the semantics are clear and semantic understanding is formed. The extracted words are named according to the predefined attribute naming rules and are not repeated.

[0031] Step 109: Perform logical judgment on the target text to identify the logical semantic word, that is, identify the condition word and the execution word in the target text, for example, identify “if...just...” in the target text.

[0032] Step 111: Perform a word cut on the target text, identify the text between the condition word and the execution word as the conditional structure text, and recognize the text after the execution word as the execution structure text. Conditional structure text and execution knot The text is stored in the corresponding data structure, for example, using JavaScript Object

Notation (json) file format storage. Conditional structure text and execution structure text can be stored in a relatively independent data structure. The sensing condition and the execution command are analyzed, that is, the condition word is recognized in the condition structure text, the production sensing condition is based on the logic analysis; the execution command word is recognized in the execution structure text, and the production execution command is based on the logic analysis.

[0033] Step 113: Execute the execution command with the precondition as a precondition, and when the execution command is triggered, execute the associated execution object.

[0034] In an embodiment of the present invention, before performing step 103, the method further includes: performing face recognition according to the image signal, and determining identity information of the user according to the recognition result. On the one hand, the illegal user can be prevented from using the smart terminal. On the other hand, the smart terminal can also provide the personalized service for the user according to the user identity information, thereby improving the use experience of the smart terminal.

[0035] As a specific example 1, when the smart terminal is a robot, the smart terminal said: "If I see the TV on the sofa and immediately hit the TV in front of the TV," the TV can be extracted. "And "TV is snoring" as "smart home" scene. Identify the condition word and the execution word as "if...just...". The perceptual condition is extracted from the conditional structure text: the location is "on the sofa", the perceptual action is "see", the character is "I", or the "sofa" and "I" are used as the conditional words. Execute the execution command in the execution structure text, the location is "in front of the TV", and the action is "TV snoring".

[0036] As a specific example 2, when speaking to the smart terminal: "If you arrive at the father to remind him to take the medicine at 11:00," you can extract the "remind him to take medicine" as the "life reminder" scene. Identify the condition word and the execution word as "if...just...". The perceptual condition "11:00" is extracted from the conditional structure text. Extract the execution command "to Dad" and "Remind him to take medicine" in the execution structure text.

[0037] As a specific example 3, when speaking to the smart terminal: "If you see someone you don't know, immediately notify me", you can extract "people you don't know" and "notify me" as "home security". Scenes. Identify the condition word and the execution word as "if...just...". In the conditional structure text, the perceptual condition "see the person who does not know" is extracted. Extract the execution command "Notify Me" in the execution structure text.

[0038] An intelligent terminal control method based on semantic analysis in an embodiment of the present invention provides an IoT intelligent terminal, for example, a semantic understanding customization in the field of service robots, and a structured semantic understanding through a concise logical semantic word. This method adds logic to the existing extraction semantics. Applied to intelligent end In the specific scene, the accuracy and efficiency of semantic understanding can be improved.

2 is a structural block diagram of a semantic terminal-based intelligent terminal control system according to an embodiment of the present invention. As shown in FIG. 2, the intelligent terminal control system 200 includes: a receiving module 210, a voice/image recognition module 220, a storage module 230, a semantic recognition module 260, an analysis module 270, an execution module 280, and a semantic knowledge library 290.

[0040] Specifically, the receiving module 210 is configured to receive an input signal, where the input signal includes a voice signal or an image signal input by a user. The speech/image recognition module 220 is for recognizing a speech signal or an image signal as text and storing it in the target text. The storage module 230 stores the target text. Semantic recognition module 260 is used to traverse the target text and identify the keywords to obtain corresponding semantic information. The analysis module 270 is configured to identify the sentence of the condition word and the execution word in the target text and recognize the sensing condition and the execution command. Execution module 280 is used to execute the associated execution object according to the execution command.

[0041] It should be noted that the specific implementation manner of the semantic terminal-based intelligent terminal control system of the embodiment of the present invention is similar to the specific implementation manner of the semantic terminal-based intelligent terminal control method according to the embodiment of the present invention. For details, refer to the method part. The description, in order to reduce redundancy, will not be described here.

Further, the present invention relates to an intelligent terminal, including the semantic analysis-based intelligent terminal control system described in the above embodiments. Other configurations and functions of the smart terminal according to the embodiments of the present invention are known to those skilled in the art, and in order to reduce redundancy, no further details are provided herein.

[0043] Please continue to refer to FIG. 3. In another embodiment, a simple semantic analysis-based intelligent terminal control method includes:

[0044] Step 301: Read the semantic knowledge base.

[0045] Step 303: Perform semantic analysis to form a semantic understanding.

[0046] Step 305: Determine whether the scene is recognized. If "Yes", go to step 307, if "No", then end [0047] Step 307: Perform specific functions.

[0048] In this embodiment, the method implements a logic and a judgment, for example, if a "if" condition is satisfied, a specific command word in "just" is executed, or if a keyword appears in "if", Execute the "implementation" function word.

[0049] In the present embodiment, the intelligent terminal control method based on semantic analysis combines unique semantic analysis technology , to achieve human-computer interaction. For specific execution functions, a method for accurately understanding natural language can be formed.

[0050] The logic and/or steps represented in the flowcharts or otherwise described herein, for example, may be considered as an ordered list of executable instructions for implementing logical functions, which may be embodied in any computer readable form. In the medium, for use by, or in conjunction with, an instruction execution system, apparatus, or device (eg, a computer-based system, a system including a processor, or other system that can fetch instructions and execute instructions from an instruction execution system, apparatus, or device) Use with systems, devices, or equipment. For the purposes of this specification, a "computer readable medium" can be any apparatus that can contain, store, communicate, propagate, or transport the program for use by the instruction execution system, apparatus, or device, or in conjunction with the instruction execution system, apparatus, or device. More specific examples (non-exhaustive list) of computer readable media include the following: electrical connections (electronic devices) having one or more wires, portable computer disk cartridges (magnetic devices), random access memory (RAM), Read only memory (ROM), erasable editable read only memory (EPROM or flash memory), fiber optic devices, and portable compact disk read only memory (CDROM). In addition, the computer readable medium may even be a paper or other suitable medium on which the program can be printed, as it may be optically scanned, for example by paper or other medium, followed by editing, interpretation or other suitable The method proceeds to obtain the program electronically and then store it in computer memory.

[0051] It should be understood that portions of the invention may be implemented in hardware, software, firmware or a combination thereof. In the above-described embodiments, multiple steps or methods may be implemented in software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented with any one or combination of the following techniques well known in the art: having logic gates for implementing logic functions on data signals Discrete logic circuit, ASIC with suitable combination logic gate, programmable gate array 歹 ij (PGA), field programmable gate array (FPGA), etc.

[0052] Those skilled in the art can understand that all or part of the steps carried by the method of implementing the above embodiments may be completed by a program to instruct related hardware, and the program may be stored in a computer readable storage medium. The program is executed, including one or a combination of the steps of the method implementation

[0053] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing module, or each unit may exist physically separately, or two or more units may be integrated into one module. . The above integrated modules can be implemented in the form of hardware or in the form of software functional modules. The integrated modules, if implemented in the form of software functional modules and sold or used as stand-alone products, may also be stored in a computer readable storage medium.

The above-mentioned storage medium may be a read only memory, a magnetic disk or an optical disk or the like. Although the embodiments of the present invention have been shown and described, it is understood that the foregoing embodiments are illustrative and are not to be construed as limiting the scope of the invention. The embodiments are subject to changes, modifications, substitutions and variations.

Claims

Claim

An intelligent terminal control method based on semantic analysis, characterized in that: the method comprises the following steps:

Providing a semantic knowledge base, the semantic knowledge base comprising a logical semantic set, the logical semantic set comprising a logical semantic word, the logical semantic word comprising a condition word and an execution word;

Recognizing text by the voice signal or image signal, and storing the recognized text in a target text;

Identifying the condition word and the execution word in the target text;

And analyzing the sensing condition and executing the command according to the condition word and the execution word; and

The execution command is executed according to the sensing condition.

A semantic terminal-based intelligent terminal control method according to claim 1, wherein

The intelligent terminal control method based on semantic analysis further comprises: traversing the target text, identifying keywords to obtain corresponding semantic information.

The method further includes: cutting a word of the target text, identifying the text between the condition word and the execution word as a conditional structure text, and identifying the text after the execution word as an execution structure text; The analysis produces the perceptual condition; and analyzes the production of the execution command in the execution structure text.

The semantic analysis-based intelligent terminal control method according to claim 3, wherein: the conditional structure text and the execution structure text are respectively stored in a JavaScript Object Notation (json) file format.

The semantic analysis-based intelligent terminal control method according to claim 3, wherein: the semantic knowledge base further comprises a scene semantic set, wherein the scene semantic set comprises a perceptual function semantic set and an execution function semantic set, and the perceptual function semantic set Include at least one perceptual condition word, the execution function semantic set includes at least one execution command word, the method further comprising: identifying a perceptual condition word in the conditional structure text, producing the perceptual condition based on logic analysis; and executing the structural text in the execution The execution command word is identified, and the execution command is produced based on logic analysis. [Claim 6] The semantic analysis based intelligent terminal control method according to claim 1, wherein

The semantic knowledge base further includes an execution object database, the execution object database storing a plurality of execution objects, each execution object associated with at least one execution command word in the execution function semantic set

[Claim 7] The semantic analysis based intelligent terminal control method according to claim 5, characterized in that

: The condition word is "if" and the execution word is "yes".

[Claim 8] An intelligent terminal control system based on semantic analysis, characterized in that: the system comprises: a semantic knowledge base comprising a logical semantic set, the logical semantic set comprising logical semantic words, the logical semantic words including condition words and Execution word

a receiving module, configured to receive an input signal, where the input signal is a voice signal or an image signal; a voice/image recognition module, configured to identify the text by the voice signal or the image signal, and store the recognized text in a target text;

An analysis module, configured to identify a condition word and an execution word in the target text, and analyze the sensing condition and the execution command according to the condition word and the execution word; and an execution module, configured to execute the execution command according to the sensing condition.

[Claim 9] The semantic analysis based intelligent terminal control system according to claim 8, wherein

The analysis module is further configured to cut the target text, identify the text between the condition word and the execution word as a conditional structure text, and identify the text after the execution word as an execution structure text; The perceptual condition is produced in the text analysis; and the execution command is analyzed and produced in the execution structure text.

[Claim 10] An intelligent terminal, comprising the semantic terminal-based intelligent terminal control system according to claim 8 or 9.