A voice interaction method, device, intelligent terminal and storage medium

Abstract

The present invention discloses a voice interaction method, device, intelligent terminal and storage medium, wherein the voice interaction method comprises: obtaining a voice command of a target object; performing voice-to-text recognition on the voice command to obtain recognition content; performing multi-dimensional semantic analysis based on the recognition content to obtain semantic analysis results; and responding based on the semantic analysis results. After performing voice-to-text recognition on the user's voice, a multi-dimensional semantic analysis is performed based on the recognition content, fully considering the dimensions of the user's thinking, thereby understanding the user's true intention based on the multi-dimensional analysis, improving the accuracy of semantic analysis recognition, and providing better voice interaction for the user.

Description

Technical Field

[0001] The present invention relates to the field of speech processing and recognition technology, and in particular to a speech interaction method, device, intelligent terminal and storage medium. Background Art

[0002] At present, with the continuous development of science and technology, artificial intelligence technology is becoming more and more mature, and various smart devices are being used more and more widely. Various smart devices can be seen everywhere in people's daily lives. With the pursuit of efficient interaction by users, voice interaction, as a new and efficient interaction method, is widely used in various smart devices, such as smart speakers, smart TVs, etc. Various smart devices can meet user needs through voice interaction.

[0003] When performing voice interaction, it is necessary to recognize the user's intention based on the acquired user voice, and then respond to complete the interaction with the user. In the prior art, the acquired user voice is usually only converted into text, and then a simple semantic analysis is performed based on the converted text, and the semantic analysis result is used as the user's intention and interacts with the user. The problem with the prior art is that it cannot accurately analyze and understand the user's true intention based on simple voice-to-text conversion and semantic analysis, the semantic analysis recognition accuracy is low, and it will cause the operations performed when interacting with the user to be not what the user wants, which is not conducive to the user's voice interaction and affects the user experience.

[0004] Therefore, the existing technology still needs to be improved and developed. Summary of the invention

[0005] In view of the technical problem that the prior art only converts the acquired user voice into text, then performs a simple semantic analysis based on the converted text, uses the semantic analysis result as the user's intention and interacts with the user, and the semantic analysis recognition accuracy is low, which is not conducive to the user's voice interaction, the present invention provides a voice interaction method, device, intelligent terminal and storage medium, which can obtain the voice command of the target object; perform voice-to-text recognition on the above voice command to obtain the recognition content; perform multi-dimensional semantic analysis based on the above recognition content to obtain the semantic analysis result; and respond based on the above semantic analysis result. After performing voice-to-text recognition on the user's voice, a multi-dimensional semantic analysis is performed based on the recognition content, fully considering the dimensions of the user's thinking problem, thereby understanding the user's true intention based on the multi-dimensional analysis, improving the accuracy of semantic analysis recognition, and providing the user with better voice interaction.

[0006] In order to achieve the above technical effects, the first aspect of the present invention provides a voice interaction method, wherein the method comprises:

[0007] Obtaining voice instructions from the target object;

[0008] Perform voice and text recognition on the above voice command to obtain recognition content;

[0009] Perform multi-dimensional semantic analysis based on the above-mentioned recognition content to obtain semantic analysis results;

[0010] Make a response based on the above semantic analysis results.

[0011] Optionally, performing multi-dimensional semantic analysis based on the above-mentioned recognition content to obtain semantic analysis results includes:

[0012] Performing semantic understanding on the above-mentioned recognition content to obtain a target domain corresponding to the above-mentioned recognition content;

[0013] Obtain multi-dimensional analysis strategies corresponding to the above target areas;

[0014] Based on the above multi-dimensional analysis strategy, a multi-dimensional semantic analysis is performed on the above identified content to obtain a semantic analysis result.

[0015] Optionally, the multi-dimensional analysis strategy for obtaining the target domain includes:

[0016] Identify and obtain the identity information of the above-mentioned target object;

[0017] Based on the above identity information, a multi-dimensional analysis strategy corresponding to the above target field is obtained.

[0018] Optionally, the above method further includes:

[0019] Record the target object’s behavioral habit data;

[0020] Based on the behavioral habit data of the above target objects, a multi-dimensional analysis strategy is generated for the above target objects.

[0021] Optionally, it is characterized in that the responding based on the semantic analysis result includes:

[0022] Generate an operation instruction based on the above semantic analysis result;

[0023] Interact with the target object based on the operation instruction.

[0024] A second aspect of the present invention provides a voice interaction device, wherein the device comprises:

[0025] An instruction acquisition module, used to acquire the voice instructions of the target object;

[0026] The command recognition module is used to perform voice and text recognition on the above voice commands to obtain recognition content;

[0027] A semantic analysis module is used to perform multi-dimensional semantic analysis based on the above-mentioned recognition content and obtain semantic analysis results;

[0028] The response control module is used to respond based on the above semantic analysis results.

[0029] Optionally, the semantic analysis module includes:

[0030] A target domain acquisition unit, used to perform semantic understanding on the above-mentioned recognition content and acquire a target domain corresponding to the above-mentioned recognition content;

[0031] An analysis strategy acquisition unit, used to acquire a multi-dimensional analysis strategy corresponding to the above target field;

[0032] The multi-dimensional semantic analysis unit is used to perform multi-dimensional semantic analysis on the above-mentioned identified content based on the above-mentioned multi-dimensional analysis strategy to obtain a semantic analysis result.

[0033] Optionally, the analysis strategy acquisition unit includes:

[0034] An identity information acquisition subunit, used to identify and acquire the identity information of the target object;

[0035] The strategy acquisition subunit is used to acquire the multi-dimensional analysis strategy corresponding to the above target field based on the above identity information.

[0036] A third aspect of the present invention provides an intelligent terminal, comprising a memory, a processor, and a program stored in the memory and executable on the processor, wherein the program implements the steps of any one of the above-mentioned voice interaction methods when executed by the processor.

[0037] A fourth aspect of the present invention provides a storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the steps of any one of the above-mentioned voice interaction methods are implemented.

[0038] As can be seen from the above, the scheme of the present invention obtains the voice command of the target object; performs speech-to-text recognition on the above voice command to obtain the recognition content; performs multi-dimensional semantic analysis based on the above recognition content to obtain the semantic analysis result; and responds based on the above semantic analysis result. Since the scheme of the present invention can perform multi-dimensional semantic analysis based on the recognition content after performing speech-to-text recognition on the user's voice, it fully considers the dimensions in which the user thinks about the problem, thereby understanding the user's true intention based on the multi-dimensional analysis. Therefore, compared with the prior art that only converts the acquired user voice into text, and then performs a simple semantic analysis based on the converted text, uses the semantic analysis result as the user's intention and interacts with the user, the scheme of the present invention can improve the accuracy of semantic analysis and recognition, which is conducive to providing users with better voice interaction. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0040] Figure 1 It is a flowchart of a voice interaction method provided by an embodiment of the present invention;

[0041] Figure 2 Embodiment of the present invention Figure 1 Specific flow diagram of step S300;

[0042] Figure 3 Embodiment of the present invention Figure 2 Specific flow diagram of step S302;

[0043] Figure 4 is a flow chart of another voice interaction method provided by an embodiment of the present invention;

[0044] Figure 5 Embodiment of the present invention Figure 1 Specific flow diagram of step S400;

[0045] Figure 6 is a structural schematic diagram of a voice interaction device provided by an embodiment of the present invention;

[0046] Figure 7 Embodiment of the present invention Figure 6 A schematic diagram of the specific structure of the semantic analysis module 630;

[0047] Figure 8 Embodiment of the present invention Figure 7 A schematic diagram of the specific structure of the analysis strategy acquisition unit 632;

[0048] Fig. 9 It is a block diagram of the internal structure principle of an intelligent terminal provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0049] In the following description, specific details such as specific system structures, technologies, etc. are provided for the purpose of illustration rather than limitation, so as to provide a thorough understanding of the embodiments of the present invention. However, it should be clear to those skilled in the art that the present invention may be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to prevent unnecessary details from obstructing the description of the present invention.

[0050] It should be understood that when used in this specification and the appended claims, the term "comprising" indicates the presence of described features, integers, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or groups thereof.

[0051] It should also be understood that the terms used in the present specification are only for the purpose of describing specific embodiments and are not intended to limit the present invention. As used in the present specification and the appended claims, unless the context clearly indicates otherwise, the singular forms "a", "an" and "the" are intended to include plural forms.

[0052] It should be further understood that the term "and / or" used in the present description and the appended claims refers to and includes any and all possible combinations of one or more of the associated listed items.

[0053] As used in this specification and the appended claims, the term "if" may be interpreted as "when" or "upon" or "in response to determining" or "in response to detecting," depending on the context. Similarly, the phrases "if it is determined" or "if [described condition or event] is detected" may be interpreted as meaning "upon determination" or "in response to determining" or "upon detection of [described condition or event]" or "in response to detecting [described condition or event]," depending on the context.

[0054] The following is a clear and complete description of the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0055] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0056] With the improvement of living standards and the development of science and technology, smart devices can be seen everywhere in people's lives. At present, most smart devices (such as smart speakers, smart TVs, smart air conditioners, smart refrigerators, etc.) support voice interaction to facilitate user operation. With the development of artificial intelligence technology and Internet of Things technology, voice interaction in the field of artificial intelligence Internet of Things (AIoT) has also received widespread attention. When performing voice interaction, it is necessary to identify the user's intention based on the acquired user voice, and then respond to complete the interaction with the user. The traditional voice interaction method in the prior art usually only converts the acquired user voice into text, and then performs a simple semantic analysis based on the converted text, and uses the semantic analysis result as the user's intention and interacts with the user. The problem with the prior art is that the user's true intention cannot be accurately analyzed and understood based on simple voice-to-text conversion and semantic analysis, and some "irrelevant" execution results often appear, which is not conducive to the user's voice interaction and greatly affects the user's operating experience. Therefore, when performing voice interaction, there is a need for a method that can improve the accuracy of user voice semantics and intention recognition.

[0057] In order to solve the problems existing in the prior art, the present invention provides a voice interaction method. In an embodiment of the present invention, a voice command of a target object is obtained; voice and text recognition is performed on the voice command to obtain recognition content; multi-dimensional semantic analysis is performed based on the recognition content to obtain semantic analysis results; and a response is made based on the semantic analysis results. After voice and text recognition is performed on the user's voice, a multi-dimensional semantic analysis is performed based on the recognition content, fully considering the dimensions of the user's thinking, thereby understanding the user's true intention based on the multi-dimensional analysis, improving the accuracy of semantic analysis recognition, and providing users with better voice interaction.

[0058] Exemplary Methods

[0059] like Figure 1 As shown, an embodiment of the present invention provides a voice interaction method, the method comprising the following steps:

[0060] Step S100, obtaining the voice command of the target object.

[0061] Among them, the above-mentioned target object is the user who needs to perform voice interaction. Optionally, the above-mentioned target object can be a specific user or all users who may issue voice commands. For example, for a certain smart device, the voiceprints of some specific users can be recorded to limit only these users to voice interact with the smart device, or there can be no restriction so that all users can voice interact with the smart device. Furthermore, restrictions can also be set for some functions of the smart device, and only some specific users can use the corresponding functions. For example, restrictions are set on the download function of the smart speaker, and all users who may issue voice commands can switch songs, but only specific users (such as administrators) can download songs. In this way, the management of voice interactions of different users is achieved and the user experience is improved.

[0062] Step S200, performing voice and text recognition on the voice command to obtain recognition content.

[0063] The above-mentioned recognition content includes the text corresponding to the user's voice command. The user's voice command is converted into text content to facilitate further processing and recognition.

[0064] Step S300: Perform multi-dimensional semantic analysis based on the above-mentioned recognition content to obtain semantic analysis results.

[0065] Among them, the analysis dimensions in the above-mentioned multi-dimensional semantic analysis include one or more of time, environment, festivals, interpersonal relationships, smart home data, weather, target object habits, and target object behaviors.

[0066] Optionally, the dimensions in the above multi-dimensional semantic analysis may also include other dimensions, such as the user's schedule, etc., which are not specifically limited here. By considering as many dimensions as possible for semantic analysis, the way and dimensions of the user's thinking can be simulated, so as to more accurately understand the user's intention and provide the user with a better voice interaction experience.

[0067] Step S400: respond based on the above semantic analysis result.

[0068] Optionally, the above response may include answering the user, controlling the status of other smart devices, arranging a trip for the user, etc., which is not specifically limited here.

[0069] As can be seen from the above, the voice interaction method provided by the embodiment of the present invention obtains the voice command of the target object; performs voice-to-text recognition on the above voice command to obtain the recognition content; performs multi-dimensional semantic analysis based on the above recognition content to obtain the semantic analysis result; and responds based on the above semantic analysis result. Since the scheme of the present invention can perform multi-dimensional semantic analysis based on the recognition content after performing voice-to-text recognition on the user's voice, it fully considers the dimensions of the user's thinking about the problem, thereby understanding the user's true intention based on the multi-dimensional analysis. Therefore, compared with the prior art that only converts the acquired user voice into text, and then performs a simple semantic analysis based on the converted text, uses the semantic analysis result as the user's intention and interacts with the user, the scheme of the present invention can improve the accuracy of semantic analysis recognition, which is conducive to providing users with better voice interaction.

[0070] Optional, such as Figure 2 As shown, in this embodiment, the above step S300 includes:

[0071] Step S301 , semantic understanding is performed on the above-mentioned recognition content to obtain a target domain corresponding to the above-mentioned recognition content.

[0072] Among them, the above-mentioned target field is the field corresponding to the identified text recognition content. The above-mentioned target field and the corresponding relationship between the above-mentioned target field and the text recognition content can be obtained by pre-setting based on user habits, obtained by big data analysis based on artificial intelligence, or obtained by corresponding user-defined adjustment settings, without specific restrictions here. Optionally, the above-mentioned target field can include food, housing, travel, clothing, etc., and there can also be other fields, without specific restrictions here.

[0073] Step S302, obtaining a multi-dimensional analysis strategy corresponding to the target field.

[0074] Among them, the above-mentioned multi-dimensional strategies can be stored locally or in the cloud in the form of a strategy table. Specifically, for different target areas, the dimensions considered and the corresponding multi-dimensional analysis strategies are different. For example, when the target area is "eating", the dimensions considered may include time, user's eating habits, remaining food in the refrigerator, etc.; when the target area is "wearing", the dimensions considered may include user gender, age, weather, temperature, user preferences, etc. In this way, for different problems, multi-dimensional considerations are made in the user's way of thinking, the user's true intentions are comprehensively understood, the accuracy of semantic analysis and recognition is improved, and a better voice interaction experience is provided for users.

[0075] Optionally, different multi-dimensional analysis strategies can be applied to the same target domain. For example, when a user says "I'm hungry" at two different times, at noon and at night, both of them correspond to the domain of "eating". However, due to the different times, different multi-dimensional analysis strategies can be applied to the user's different needs for breakfast and lunch.

[0076] Step S303: Perform multi-dimensional semantic analysis on the identified content based on the multi-dimensional analysis strategy to obtain a semantic analysis result.

[0077] In this way, user needs are fully taken into consideration and the accuracy of semantic analysis and recognition is improved.

[0078] Optional, such as Figure 3 As shown, in this embodiment, the above step S302 includes:

[0079] Step S3021, identifying and obtaining the identity information of the target object.

[0080] Step S3022: acquiring a multi-dimensional analysis strategy corresponding to the target domain based on the identity information.

[0081] Optionally, the user's identity information may be obtained through voiceprint recognition, or through face recognition of the user. There may also be other identification methods, which are not specifically limited here.

[0082] In this embodiment, different users can correspond to different multi-dimensional analysis strategies, thereby providing customized voice interaction for different users and improving user experience. Specifically, the multi-dimensional analysis strategies corresponding to different users can be customized by the users themselves, or can be intelligently generated through recorded user habit data.

[0083] Optionally, when the corresponding user identity information cannot be obtained, or the user has not set the corresponding multi-dimensional analysis strategy and the number of recorded user habit data is lower than the preset strategy data threshold, and the corresponding multi-dimensional analysis strategy cannot be intelligently generated for the user, the corresponding default strategy can be obtained for the user. Among them, the above-mentioned strategy data threshold is a pre-set critical data threshold for intelligently generating the corresponding multi-dimensional analysis strategy for the user. Specifically, the basic attributes of the user can be first judged by the user's voiceprint and the user's environment, such as judging that the user is a 25-year-old office white-collar female, and further matching the above-mentioned user basic attributes to obtain the living habits of people with the same attributes in the area in the cloud, and multi-dimensional semantic analysis is performed based on the living habits of the above-mentioned people with the same attributes and some of the user's setting data to further complete the interaction with the user. For example, when the voice command of the user obtained is "I'm hungry", the analysis output may be set based on the living habits of people with the same attributes and the user's schedule: "You have an appointment with someone at noon to eat at a certain restaurant. Do you need to make a reservation in advance?", where a person and a restaurant should correspond to a specific restaurant and a person's name, which is only an example here.

[0084] Optional, such as Figure 4 As shown, in this embodiment, the voice interaction method further includes:

[0085] Step A100, recording the behavior habit data of the target object.

[0086] Specifically, the user's behavior habit data can be recorded in real time through all associated smart devices. Optionally, when there are multiple users, the behavior habit data of different users can be recorded in the database corresponding to the user to avoid confusion of the behavior habit data of different users.

[0087] Step A200, generating a multi-dimensional analysis strategy for the target object based on the behavioral habit data of the target object.

[0088] Specifically, when the recorded user behavior habit data reaches the above-mentioned preset strategy data threshold, a multi-dimensional analysis strategy can be generated for the user. Optionally, when the user already has a corresponding multi-dimensional analysis strategy and the user's behavior habits change, the above multi-dimensional analysis strategy can also be updated accordingly, thereby providing the user with a better interactive experience.

[0089] Optional, such as Figure 5 As shown, in this embodiment, the above step S400 includes:

[0090] Step S401: Generate an operation instruction based on the above semantic analysis result.

[0091] Step S402: interact with the target object based on the operation instruction.

[0092] Specifically, after obtaining the above semantic analysis results and clarifying the user's true intention, an operation instruction can be generated, and the target object can be interacted with based on the operation instruction. The above operation instruction can include voice reply instructions and control instructions, and the above operation instruction can be used to operate and control all associated smart devices to enhance the user interaction experience.

[0093] In this embodiment, the above-mentioned voice interaction process is specifically illustrated by examples. For example, in an application scenario, a voice command "I am hungry" issued by a user is received, and the above-mentioned voice command is recognized by voice and text, and the recognition content "I am hungry" is obtained. A simple semantic understanding is performed on the above-mentioned recognition content, and it is obtained that the above-mentioned recognition content is related to the field of "eating", so a multi-dimensional analysis strategy related to "eating" is obtained, and a multi-dimensional semantic analysis is performed based on the multi-dimensional strategy. Specifically, the following strategy steps are executed: Strategy step a, obtain the current time, specifically, determine when the current user is hungry, whether it is morning, noon, evening, morning, afternoon, late night or other time periods, so as to further determine whether it is meal time or not; Strategy step b, obtain user habits based on the above current time, specifically, if it is determined that the time when the user is hungry is meal time in the morning, then obtain the user habits of the corresponding time, and determine whether the user likes to cook breakfast at home, take breakfast from the refrigerator or eat breakfast nearby; Strategy step c, based on the above user habits, determine the user intention as a result of semantic analysis. Specifically, if the user's habit is to take breakfast from the refrigerator, then the user's intention may be to obtain a list of breakfast foods in the refrigerator and recommend them to the user. If the user's habit is to eat breakfast nearby, then the user's intention may be to obtain today's weather and recommend a corresponding breakfast shop.

[0094] Optionally, if the current time is noon, that is, the user is hungry at noon, then the user's corresponding lunch habits should be obtained in the specific strategy steps. The specific process is similar to the above process and will not be repeated here.

[0095] Furthermore, after performing the above-mentioned multi-dimensional semantic analysis, a response can be made based on the above-mentioned user intention. For example, when the user intends to obtain a list of breakfast foods in the refrigerator and recommend them to the user, the smart refrigerator can be accessed to obtain the list of breakfast foods, and a voice reply to the user can be made: "There is still some of your favorite food in the refrigerator, please eat breakfast on time". If the user's refrigerator does not have the breakfast food that the user often eats, the breakfast food list is empty, and the user can also be reminded to buy it and the purchase list can be recommended to the user. When the user intends to obtain today's weather and recommend the corresponding breakfast shop, the weather information can be accessed. If it is sunny, the user can be replied to by voice: "The weather is sunny, you can go out for breakfast". Breakfast shops are recommended to users, and navigation is provided to users based on the data of surrounding breakfast shops. In this way, the user's true intention is understood based on multi-dimensional analysis, the accuracy of semantic analysis recognition is improved, and a better voice interaction experience is provided to users.

[0096] In another application scenario, a voice command "I feel stuffy" issued by a user is received, and the voice command is recognized by voice and text, and the recognition content "I feel stuffy" is obtained. A simple semantic understanding is performed on the above recognition content, and the above recognition content is obtained to be related to the "perception" field, so a multi-dimensional analysis strategy related to "perception" is obtained, and a multi-dimensional semantic analysis is performed and a response is made based on the multi-dimensional strategy. Specifically, the following strategy steps are executed and responded to at the same time, and interaction is performed: Strategy step a, analyze from the physical layer, specifically, obtain weather information and temperature information, confirm whether the doors and windows are opened for ventilation, whether the air conditioning temperature is appropriate, etc. If the doors and windows and air conditioning settings are not appropriate, an adjustment control instruction is generated, and the above troubleshooting results and adjustment intentions are fed back to the user, and the execution is executed after the user confirms the execution, and the comfort level is confirmed after the above adjustment control instruction is executed for a certain period of time. Strategy step b, analyze from the psychological level, specifically, the family environment atmosphere can be adjusted, such as adjusting the lights and music, and there can be other adjustment methods, which are not specifically limited here. Strategy step c, analyze from the social level, specifically, you can obtain the user's social information, analyze and give suggestions based on the above social information, for example, you can reply to the user by voice: "Today is a holiday, you can go out to eat with someone" or "I found someone somewhere, do you want me to make an appointment for you", and you can also distract the user's attention through real-time social dynamics, etc. Strategy step d, analyze from the physiological level, specifically, you can use the temperature test of the camera and the user's physiological characteristics through the smart bracelet to detect whether the user is in a pathological state. If so, remind the user to seek medical treatment, query the case record and medication record, and remind the user to take the medicine if there is a medicine that meets the characteristics. If necessary, you can start the emergency medical system.

[0097] Exemplary Devices

[0098] like Figure 6 As shown in , corresponding to the above-mentioned voice interaction method, an embodiment of the present invention further provides a voice interaction device, and the above-mentioned voice interaction device includes:

[0099] The instruction acquisition module 610 is used to acquire the voice instruction of the target object.

[0100] Among them, the above-mentioned target object is the user who needs to perform voice interaction. Optionally, the above-mentioned target object can be a specific user or all users who may issue voice commands. For example, for a certain smart device, the voiceprints of some specific users can be recorded to limit only these users to voice interact with the smart device, or there can be no restriction so that all users can voice interact with the smart device. Furthermore, restrictions can also be set for some functions of the smart device, and only some specific users can use the corresponding functions. For example, restrictions are set on the download function of the smart speaker, and all users who may issue voice commands can switch songs, but only specific users (such as administrators) can download songs. In this way, the management of voice interactions of different users is achieved and the user experience is improved.

[0101] The instruction recognition module 620 is used to perform voice and text recognition on the above voice instructions to obtain recognition content.

[0102] The above-mentioned recognition content includes the text corresponding to the user's voice command. The user's voice command is converted into text content to facilitate further processing and recognition.

[0103] The semantic analysis module 630 is used to perform multi-dimensional semantic analysis based on the above-mentioned recognition content to obtain semantic analysis results.

[0104] Among them, the analysis dimensions in the above-mentioned multi-dimensional semantic analysis include one or more of time, environment, festivals, interpersonal relationships, smart home data, weather, target object habits, and target object behaviors.

[0105] Optionally, the dimensions in the above multi-dimensional semantic analysis may also include other dimensions, such as the user's schedule, etc., which are not specifically limited here. By considering as many dimensions as possible for semantic analysis, the way and dimensions of the user's thinking can be simulated, so as to more accurately understand the user's intention and provide the user with a better voice interaction experience.

[0106] The response control module 640 is used to make a response based on the above semantic analysis result.

[0107] Optionally, the above response may include answering the user, controlling the status of other smart devices, arranging a trip for the user, etc., which is not specifically limited here.

[0108] As can be seen from the above, the voice interaction device provided by the embodiment of the present invention obtains the voice command of the target object through the instruction acquisition module 610; performs voice-to-text recognition on the above voice command through the instruction recognition module 620 to obtain the recognition content; performs multi-dimensional semantic analysis based on the above recognition content through the semantic analysis module 630 to obtain the semantic analysis result; and responds based on the above semantic analysis result through the response control module 640. Since the scheme of the present invention can perform multi-dimensional semantic analysis based on the recognition content after performing voice-to-text recognition on the user's voice, it fully considers the dimensions of the user's thinking problem, thereby understanding the user's true intention based on the multi-dimensional analysis. Therefore, compared with the prior art that only converts the acquired user voice into text, and then performs a simple semantic analysis based on the converted text, uses the semantic analysis result as the user's intention and interacts with the user, the scheme of the present invention can improve the accuracy of semantic analysis recognition, which is conducive to providing users with better voice interaction.

[0109] Optional, such as Figure 7 As shown, in this embodiment, the semantic analysis module 630 includes:

[0110] The target domain acquisition unit 631 is used to perform semantic understanding on the above-mentioned recognition content and acquire the target domain corresponding to the above-mentioned recognition content.

[0111] Among them, the above-mentioned target field is the field corresponding to the identified text recognition content. The above-mentioned target field and the corresponding relationship between the above-mentioned target field and the text recognition content can be obtained by pre-setting based on user habits, obtained by big data analysis based on artificial intelligence, or obtained by corresponding user-defined adjustment settings, without specific restrictions here. Optionally, the above-mentioned target field can include food, housing, travel, clothing, etc., and there can also be other fields, without specific restrictions here.

[0112] The analysis strategy acquisition unit 632 is used to acquire the multi-dimensional analysis strategy corresponding to the above target field.

[0113] Among them, the above-mentioned multi-dimensional strategies can be stored locally or in the cloud in the form of a strategy table. Specifically, for different target areas, the dimensions considered and the corresponding multi-dimensional analysis strategies are different. For example, when the target area is "eating", the dimensions considered may include time, user's eating habits, remaining food in the refrigerator, etc.; when the target area is "wearing", the dimensions considered may include user gender, age, weather, temperature, user preferences, etc. In this way, for different problems, multi-dimensional considerations are made in the user's way of thinking, the user's true intentions are comprehensively understood, the accuracy of semantic analysis and recognition is improved, and a better voice interaction experience is provided for users.

[0114] Optionally, different multi-dimensional analysis strategies can be applied to the same target domain. For example, when a user says "I'm hungry" at two different times, at noon and at night, both of them correspond to the domain of "eating". However, due to the different times, different multi-dimensional analysis strategies can be applied to the user's different needs for breakfast and lunch.

[0115] The multi-dimensional semantic analysis unit 633 is used to perform multi-dimensional semantic analysis on the above-mentioned recognition content based on the above-mentioned multi-dimensional analysis strategy to obtain a semantic analysis result.

[0116] In this way, user needs are fully taken into consideration and the accuracy of semantic analysis and recognition is improved.

[0117] Optional, such as Figure 8 As shown, in this embodiment, the analysis strategy acquisition unit 632 includes:

[0118] The identity information acquisition subunit 6321 is used to identify and acquire the identity information of the target object.

[0119] The strategy acquisition subunit 6322 is used to acquire the multi-dimensional analysis strategy corresponding to the above-mentioned target field based on the above-mentioned identity information.

[0120] Optionally, the user's identity information may be obtained through voiceprint recognition, or through face recognition of the user. There may also be other identification methods, which are not specifically limited here.

[0121] In this embodiment, different users can correspond to different multi-dimensional analysis strategies, thereby providing customized voice interaction for different users and improving user experience. Specifically, the multi-dimensional analysis strategies corresponding to different users can be customized by the users themselves, or can be intelligently generated through recorded user habit data.

[0122] Optionally, when the corresponding user identity information cannot be obtained, or the user has not set the corresponding multi-dimensional analysis strategy and the number of recorded user habit data is lower than the preset strategy data threshold, and the corresponding multi-dimensional analysis strategy cannot be intelligently generated for the user, the corresponding default strategy can be obtained for the user. Among them, the above-mentioned strategy data threshold is a pre-set critical data threshold for intelligently generating the corresponding multi-dimensional analysis strategy for the user. Specifically, the basic attributes of the user can be first judged by the user's voiceprint and the user's environment, such as judging that the user is a 25-year-old office white-collar female, and further matching the above-mentioned user basic attributes to obtain the living habits of people with the same attributes in the area in the cloud, and multi-dimensional semantic analysis is performed based on the living habits of the above-mentioned people with the same attributes and some of the user's setting data to further complete the interaction with the user. For example, when the voice command of the user obtained is "I'm hungry", the analysis output may be set based on the living habits of people with the same attributes and the user's schedule: "You have an appointment with someone at noon to eat at a certain restaurant. Do you need to make a reservation in advance?", where a person and a restaurant should correspond to a specific restaurant and a person's name, which is only an example here.

[0123] Optionally, the above-mentioned voice interaction device also includes: a habit data recording module, which is used to record the behavioral habit data of the target object; and a strategy generation module, which is used to generate a multi-dimensional analysis strategy for the above-mentioned target object based on the behavioral habit data of the above-mentioned target object.

[0124] Specifically, the user's behavior habit data can be recorded in real time through all associated smart devices. Optionally, when there are multiple users, the behavior habit data of different users can be recorded in the database corresponding to the user to avoid confusion of the behavior habit data of different users.

[0125] Specifically, when the recorded user behavior habit data reaches the above-mentioned preset strategy data threshold, a multi-dimensional analysis strategy can be generated for the user. Optionally, when the user already has a corresponding multi-dimensional analysis strategy and the user's behavior habits change, the above multi-dimensional analysis strategy can also be updated accordingly, thereby providing the user with a better interactive experience.

[0126] Optionally, the response control module is specifically used to: generate an operation instruction based on the semantic analysis result; and interact with the target object based on the operation instruction.

[0127] Specifically, after obtaining the above semantic analysis results and clarifying the user's true intention, an operation instruction can be generated, and the target object can be interacted with based on the operation instruction. The above operation instruction can include voice reply instructions and control instructions, and the above operation instruction can be used to operate and control all associated smart devices to enhance the user interaction experience.

[0128] In this embodiment, the above-mentioned voice interaction process is specifically illustrated by examples. For example, in an application scenario, a voice command "I am hungry" issued by a user is received, and the above-mentioned voice command is recognized by voice and text, and the recognition content "I am hungry" is obtained. A simple semantic understanding is performed on the above-mentioned recognition content, and it is obtained that the above-mentioned recognition content is related to the field of "eating", so a multi-dimensional analysis strategy related to "eating" is obtained, and a multi-dimensional semantic analysis is performed based on the multi-dimensional strategy. Specifically, the following strategy steps are executed: Strategy step a, obtain the current time, specifically, determine when the current user is hungry, whether it is morning, noon, evening, morning, afternoon, late night or other time periods, so as to further determine whether it is meal time or not; Strategy step b, obtain user habits based on the above current time, specifically, if it is determined that the time when the user is hungry is meal time in the morning, then obtain the user habits of the corresponding time, and determine whether the user likes to cook breakfast at home, take breakfast from the refrigerator or eat breakfast nearby; Strategy step c, based on the above user habits, determine the user intention as a result of semantic analysis. Specifically, if the user's habit is to take breakfast from the refrigerator, then the user's intention may be to obtain a list of breakfast foods in the refrigerator and recommend them to the user. If the user's habit is to eat breakfast nearby, then the user's intention may be to obtain today's weather and recommend a corresponding breakfast shop.

[0129] Optionally, if the current time is noon, that is, the user is hungry at noon, then the user's corresponding lunch habits should be obtained in the specific strategy steps. The specific process is similar to the above process and will not be repeated here.

[0130] Furthermore, after performing the above-mentioned multi-dimensional semantic analysis, a response can be made based on the above-mentioned user intention. For example, when the user intends to obtain a list of breakfast foods in the refrigerator and recommend them to the user, the smart refrigerator can be accessed to obtain the list of breakfast foods, and a voice reply to the user can be made: "There is still some of your favorite food in the refrigerator, please eat breakfast on time". If the user's refrigerator does not have the breakfast food that the user often eats, the breakfast food list is empty, and the user can also be reminded to buy it and the purchase list can be recommended to the user. When the user intends to obtain today's weather and recommend the corresponding breakfast shop, the weather information can be accessed. If it is sunny, the user can be replied to by voice: "The weather is sunny, you can go out for breakfast". Breakfast shops are recommended to users, and navigation is provided to users based on the data of surrounding breakfast shops. In this way, the user's true intention is understood based on multi-dimensional analysis, the accuracy of semantic analysis recognition is improved, and a better voice interaction experience is provided to users.

[0131] In another application scenario, a voice command "I feel stuffy" issued by a user is received, and the voice command is recognized by voice and text, and the recognition content "I feel stuffy" is obtained. A simple semantic understanding is performed on the above recognition content, and the above recognition content is obtained to be related to the "perception" field, so a multi-dimensional analysis strategy related to "perception" is obtained, and a multi-dimensional semantic analysis is performed and a response is made based on the multi-dimensional strategy. Specifically, the following strategy steps are executed and responded to at the same time, and interaction is performed: Strategy step a, analyze from the physical layer, specifically, obtain weather information and temperature information, confirm whether the doors and windows are opened for ventilation, whether the air conditioning temperature is appropriate, etc. If the doors and windows and air conditioning settings are not appropriate, an adjustment control instruction is generated, and the above troubleshooting results and adjustment intentions are fed back to the user, and the execution is executed after the user confirms the execution, and the comfort level is confirmed after the above adjustment control instruction is executed for a certain period of time. Strategy step b, analyze from the psychological level, specifically, the family environment atmosphere can be adjusted, such as adjusting the lights and music, and there can be other adjustment methods, which are not specifically limited here. Strategy step c, analyze from the social level, specifically, you can obtain the user's social information, analyze and give suggestions based on the above social information, for example, you can reply to the user by voice: "Today is a holiday, you can go out to eat with someone" or "I found someone somewhere, do you want me to make an appointment for you", and you can also distract the user's attention through real-time social dynamics, etc. Strategy step d, analyze from the physiological level, specifically, you can use the temperature test of the camera and the user's physiological characteristics through the smart bracelet to detect whether the user is in a pathological state. If so, remind the user to seek medical treatment, query the case record and medication record, and remind the user to take the medicine if there is a medicine that meets the characteristics. If necessary, you can start the emergency medical system.

[0132] Based on the above embodiments, the present invention further provides an intelligent terminal, whose principle block diagram can be shown as follows: Fig. 9 As shown. The above-mentioned intelligent terminal includes a processor, a memory, a network interface and a display screen connected through a system bus. Among them, the processor of the intelligent terminal is used to provide computing and control capabilities. The memory of the intelligent terminal includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The network interface of the intelligent terminal is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, the steps of any one of the above-mentioned voice interaction methods are implemented. The display screen of the intelligent terminal can be a liquid crystal display screen or an electronic ink display screen.

[0133] Those skilled in the art will understand that Fig. 9The principle block diagram shown in the figure is only a block diagram of a partial structure related to the scheme of the present invention, and does not constitute a limitation on the smart terminal to which the scheme of the present invention is applied. The specific smart terminal may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0134] In one embodiment, a smart terminal is provided, comprising a memory, a processor, and a program stored in the memory and executable on the processor, wherein the program performs the following operation instructions when executed by the processor:

[0135] Obtaining voice instructions from the target object;

[0136] Perform voice and text recognition on the above voice command to obtain recognition content;

[0137] Perform multi-dimensional semantic analysis based on the above-mentioned recognition content to obtain semantic analysis results;

[0138] Make a response based on the above semantic analysis results.

[0139] An embodiment of the present invention further provides a storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of any voice interaction method provided in the embodiment of the present invention are implemented.

[0140] It should be understood that the serial numbers of the steps in the above embodiments do not imply a sequence of execution, and the execution sequence of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0141] Those skilled in the art can clearly understand that, for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example for illustration. In practical applications, the above-mentioned function allocation can be completed by different functional units and modules as needed, that is, the internal structure of the above-mentioned device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiment can be integrated in a processing unit, or each unit can exist physically separately, or two or more units can be integrated in one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of the present invention. The specific working process of the units and modules in the above-mentioned system can refer to the corresponding process in the aforementioned method embodiment, which will not be repeated here.

[0142] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described or recorded in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0143] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0144] In the embodiments provided by the present invention, it should be understood that the disclosed apparatus / terminal equipment and method can be implemented in other ways. For example, the apparatus / terminal equipment embodiments described above are only illustrative, for example, the division of the above modules or units is only a logical function division, and in actual implementation, other division methods can be used, for example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed.

[0145] If the above-mentioned integrated module / unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on such an understanding, the present invention implements all or part of the processes in the above-mentioned embodiment method, and can also be completed by instructing the relevant hardware through a computer program. The above-mentioned computer program can be stored in a computer-readable storage medium, and the computer program can implement the steps of the above-mentioned various method embodiments when executed by the processor. Among them, the above-mentioned computer program includes computer program code, and the above-mentioned computer program code can be in source code form, object code form, executable file or some intermediate form. The above-mentioned computer-readable medium may include: any entity or device capable of carrying the above-mentioned computer program code, recording medium, U disk, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), electric carrier signal, telecommunication signal and software distribution medium. It should be noted that the content contained in the above-mentioned computer-readable storage medium can be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction.

[0146] The embodiments described above are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, it should be understood by those skilled in the art that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features therein may be replaced by equivalents. However, these modifications or replacements do not deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should be included in the protection scope of the present invention.

Claims

1. A voice interaction method, characterized in that: The method comprises: Obtaining voice instructions from the target object; Performing voice and text recognition on the voice command to obtain recognition content; Perform multi-dimensional semantic analysis based on the identified content to obtain semantic analysis results; The performing multi-dimensional semantic analysis based on the identified content to obtain semantic analysis results includes: Performing semantic understanding on the identified content to obtain a target domain corresponding to the identified content; Obtaining a multi-dimensional analysis strategy corresponding to the target field; Performing a multi-dimensional semantic analysis on the identified content based on the multi-dimensional analysis strategy to obtain a semantic analysis result; The obtaining of the multi-dimensional analysis strategy corresponding to the target field includes: Identify and obtain the identity information of the target object; Acquire a multi-dimensional analysis strategy corresponding to the target field based on the identity information; When the identity information of the target object cannot be obtained, the basic attributes of the user are determined through the user's voiceprint and the environment in which the user is located. Based on the basic attributes, the living habits of users with the same basic attributes within a preset area are obtained through cloud matching. The multi-dimensional analysis strategy corresponding to the target area is obtained based on the living habits and part of the user's setting data; and a response is made based on the semantic analysis result; The method further comprises: Record the target object’s behavioral habit data; Based on the behavioral habit data of the target object, a multi-dimensional analysis strategy is generated for the target object; The multi-dimensional analysis strategy includes physical layer, psychological layer and social layer.

2. The voice interaction method according to claim 1, characterized in that: The responding based on the semantic analysis result includes: Generate an operation instruction based on the semantic analysis result; Interact with the target object based on the operation instruction.

3. A voice interaction device, characterized in that: The device comprises: An instruction acquisition module, used to acquire the voice instructions of the target object; An instruction recognition module is used to perform voice and text recognition on the voice instruction to obtain recognition content; A semantic analysis module, used to perform multi-dimensional semantic analysis based on the identified content and obtain semantic analysis results; The semantic analysis module includes: A target domain acquisition unit, configured to perform semantic understanding on the identified content and acquire a target domain corresponding to the identified content; An analysis strategy acquisition unit, used to acquire a multi-dimensional analysis strategy corresponding to the target field; A multi-dimensional semantic analysis unit, configured to perform a multi-dimensional semantic analysis on the identified content based on the multi-dimensional analysis strategy to obtain a semantic analysis result; The analysis strategy acquisition unit includes: An identity information acquisition subunit, used to identify and acquire the identity information of the target object; A strategy acquisition subunit, used to acquire a multi-dimensional analysis strategy corresponding to the target domain based on the identity information; When the identity information of the target object cannot be obtained, the basic attributes of the user are determined by the user's voiceprint and the environment in which the user is located. Based on the basic attributes, the living habits of users with the same basic attributes within a preset area are obtained through cloud matching. The multi-dimensional analysis strategy corresponding to the target area is obtained based on the living habits and part of the user's setting data; a response control module is used to respond based on the semantic analysis result; Record the target object’s behavioral habit data; Based on the behavioral habit data of the target object, a multi-dimensional analysis strategy is generated for the target object; The multi-dimensional analysis strategy includes physical layer, psychological layer and social layer.

4. An intelligent terminal, characterized in that: The method comprises a memory, a processor and a program stored in the memory and executable on the processor, wherein the program, when executed by the processor, implements the steps of the method according to any one of claims 1 to 2.

5. A storage medium, characterized in that: The storage medium stores a computer program, which, when executed by a processor, implements the steps of the method according to any one of claims 1 to 2.

Images