Equipment control method and device based on semantic analysis, controlled equipment and server

Abstract

The invention relates to the technical field of semantic analysis, intelligent control, artificial intelligence, man-machine interaction and the like, and provides an equipment control method and device based on semantic analysis, controlled equipment and a server. And automatically controlling the working mode of the controlled equipment according to the sense feature semantics, so that the controlled equipment acts on the equipment user in different working modes according to different sense feature semantics, and the use experience of the equipment user when the equipment user uses the controlled equipment is changed. Therefore, the association degree among the semantics, the controlled equipment and the use experience of the controlled equipment is established, the intelligent degree of equipment control is improved, and the user experience is improved.

Description

Technical Field

[0001] The present invention relates to the technical fields of semantic analysis, intelligent control, artificial intelligence, human-computer interaction, etc., and particularly relates to a device control method, device, controlled device and server based on semantic analysis. Background Art

[0002] With the rapid development of information technology, many service-oriented intelligent devices have been able to act on the user's body or living environment through different control methods to provide various personalized usage experiences. Existing intelligent devices usually adopt traditional methods such as button control, touch screen control, remote control, and mobile phone APP control to adjust the working state and functions of the device. For example, the vibration frequency of a massager can be adjusted through buttons, touch screens or remote controls, and users can also remotely control its functions through a mobile phone APP. Although these traditional control methods provide users with basic device operation functions, they still have a high dependence on user operations. Users need to perform multiple manual interactions to achieve the expected device behavior. Moreover, the existing device control methods cannot be adjusted in real time according to the user's emotional or psychological state, and cannot achieve the effect of emotional interaction, resulting in a relatively single and mechanical usage experience of intelligent devices. Therefore, there is a large room for improvement in the interactivity, personalization, and emotional response ability of existing intelligent devices. Traditional service-oriented intelligent devices, such as massagers, household appliances, and smart homes, although they have certain intelligent control functions, their ability to perceive and respond to the user's emotional state is still insufficient. Most devices still rely on direct user operations to trigger preset functions and cannot actively adjust according to semantics to allow users to obtain an immersive experience.

[0003] In summary, the existing technology has technical problems such as relying on user operations, low intelligence level, inability to actively adjust according to semantics, and inability to allow users to obtain an immersive experience. Summary of the Invention

[0004] Aiming at the deficiencies of the above-mentioned existing technology, the present invention provides a device control method, device, controlled device and server based on semantic analysis to improve the intelligence level of intelligent device control and enhance the user usage experience of intelligent devices.

[0005] In a first aspect, the present invention provides a device control method based on semantic analysis, including: Performing semantic analysis on semantic text to obtain sensory feature semantics that affect the sensory state of the device user of the controlled device; Automatically controlling the working mode of the controlled device according to the sensory feature semantics, so that the controlled device acts on the device user in different working modes according to different sensory feature semantics, and changes the usage experience when the device user uses the controlled device.

[0006] In a second aspect, the present invention provides a device control apparatus based on semantic analysis, including: A semantic analysis module, configured to perform semantic analysis on semantic texts to obtain semantic features of a feeling state of a device user affecting a controlled device; A working mode control module, configured to automatically control a working mode of the controlled device according to the semantic features of the feeling state, so that the controlled device acts on the device user in different working modes according to different semantic features of the feeling state, and change the usage experience of the device user when using the controlled device.

[0007] In a third aspect, the present invention provides a controlled device, which is controlled by using the above-mentioned device control method based on semantic analysis.

[0008] In a fourth aspect, the present invention provides a server, which runs the above-mentioned device control method based on semantic analysis to control a controlled device.

[0009] Compared with the prior art, the present invention has the following beneficial effects: The present invention provides a device control method, apparatus, controlled device and server based on semantic analysis. By performing semantic analysis on semantic texts to obtain semantic features of a feeling state of a device user affecting a controlled device, and automatically controlling a working mode of the controlled device according to the semantic features of the feeling state, so that the controlled device acts on the device user in different working modes according to different semantic features of the feeling state, and change the usage experience of the device user when using the controlled device, thereby establishing a correlation between semantics, the controlled device and the usage experience of the controlled device, improving the intelligence level of device control, and improving the user experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] The drawings described herein are used to provide a further understanding of the present invention, and constitute a part of the present invention. The illustrative embodiments and descriptions thereof of the present invention are used to explain the present invention, and do not constitute an improper limitation to the present invention. Some specific embodiments of the present invention will be described in detail hereinafter with reference to the drawings in an exemplary rather than restrictive manner. The same reference numerals in the drawings denote the same or similar components or parts. Those skilled in the art should understand that these drawings are not necessarily drawn to scale. In the drawings: Figure 1 is a schematic flowchart of a device control method based on semantic analysis according to an embodiment of the present invention; Figure 2 is a schematic architecture diagram of a device control method apparatus based on semantic analysis according to an embodiment of the present invention; Figure 3It is a schematic architecture diagram of the server in an embodiment of the present invention; Figure 4 It is a schematic architecture diagram of the controlled device communicating with the user terminal and the server in an embodiment of the present invention. Detailed implementation manners

[0013] In order to enable those skilled in the art of the present technology to better understand the solution of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0014] Embodiment 1 Refer to Figures 1 - 4 , this embodiment provides a device control method based on semantic analysis, including step S101 and step S102. Among them, step S101 and step S102 can run on the server or the user terminal. By performing semantic analysis on the semantic text, the sensory feature semantics that affect the sensory state of the device user of the controlled device are obtained, and the working mode of the controlled device is automatically controlled according to the sensory feature semantics, so that the controlled device acts on the device user in different working modes according to the differences in the sensory feature semantics, changing the usage experience of the device user when using the controlled device, thereby establishing the correlation between semantics, the controlled device, and the usage experience of the controlled device, improving the intelligence level of device control, and improving the user experience.

[0015] It should be noted that the server includes a memory, a processor, and a network interface that are communicatively connected to each other through a system bus. Among them, those skilled in the art of the present technology can understand that the server here is a device that can automatically perform numerical calculations and / or information processing according to pre-set or stored instructions, and its hardware includes, but is not limited to, a microprocessor, an application specific integrated circuit (ASIC), a field-programmable gate array (FPGA), a digital signal processor (DSP), an embedded device, etc. The server can perform human-computer interaction with users through methods such as a keyboard, a mouse, a remote control, a touchpad, or a voice control device. The memory includes at least one type of readable storage medium, and the readable storage medium includes flash memory, a hard disk, a multimedia card, a card-type memory (such as an SD or DX memory, etc.), a random access memory (RAM), a static random access memory (SRAM), a read-only memory (ROM), an electrically erasable programmable read-only memory (EEPROM), a programmable read-only memory (PROM), a magnetic memory, a magnetic disk, an optical disk, etc. In some embodiments, the memory can be an internal storage unit of the server, such as the hard disk or memory of the server. In other embodiments, the memory can also be an external storage device of the server, such as a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc. equipped on the server. Of course, the memory can also include both the internal storage unit of the server and its external storage devices.

[0016] See Figures 1 - 4 , the device control method based on semantic analysis provided in this embodiment may include the following steps: Step S101: Perform semantic analysis on the semantic text to obtain the semantic of the sensory feature that affects the sensory state of the device user of the controlled device; Step S102: Automatically control the working mode of the controlled device according to the semantic of the sensory feature, so that the controlled device acts on the device user in different working modes according to the difference of the semantic of the sensory feature, and changes the usage experience of the device user when using the controlled device.

[0017] It should be noted that traditional service-oriented intelligent devices (such as massagers, household appliances, smart homes, etc.) mainly rely on multiple manual operations by users to achieve the expected functions, and it is difficult to make real-time and active function adjustments according to the user's emotional or psychological state. In step S101, by performing semantic analysis on the semantic text, the semantic features of the sensory state of the device user affecting the controlled device are obtained, so as to actively mine the emotions, needs or psychological states in the semantic text. Step S101 identifies and analyzes the semantic information in the semantic text, greatly reducing the frequency of active operations by users and realizing the automatic capture of the user's emotional and psychological states. In addition, the traditional control method lacks emotional interaction, and the use experience is single and mechanical. There is a lack of real-time perception of the user's emotions and feelings, and it is difficult to create an immersive experience. In step S102, the working mode of the controlled device is automatically controlled according to the sensory feature semantics, that is, the sensory feature semantics obtained by semantic analysis are linked with the device working mode to realize the automatic and personalized adjustment of the device. By mapping the emotional semantic recognition result to different working modes of the controlled device (such as a massager) (reciprocating motion, rotation, vibration frequency, etc.), the user's use experience is adjusted and enhanced in real time, making the interaction mode of the device more flexible and emotionally responsive, and significantly improving the intelligence and active adjustment ability of device control. Among them, the controlled device includes a massager; the working modes of the massager include a reciprocating motion mode based on the sensory feature semantics, and / or a rotational motion mode, and / or different vibration frequency modes and / or different vibration intensity modes.

[0018] In some preferred embodiments, the semantic text is a text text or a voice text; the voice text is an audio text in a dialogue form or an audio text in a non-dialogue form; the audio text in the dialogue form is an audio text in the current dialogue form or a recorded text in the dialogue form. It should be noted that the emotional or demand information of the user often exists in various forms such as text texts or voice texts. The user may express emotions and demands through text (such as chat records, content of novel documents) or audio (such as conversations, recordings). In this embodiment, the semantic text may include various source forms. For example, it may include text texts, audio texts in the dialogue form, audio texts in the non-dialogue form, etc., so that the device control method is applicable to different input channels, thereby meeting more diverse usage scenarios (such as real-time interaction or analysis of recorded files). The voice text can come from either the audio text in the dialogue form (current dialogue or recorded text in the dialogue form) or the audio text in the non-dialogue form (such as novel recordings, etc.). Different voice scenarios will contain different contexts and emotional characteristics. In this embodiment, by classifying and defining the audio text in the dialogue form and the audio text in the non-dialogue form, the comprehensiveness in collecting and analyzing semantics is ensured. In addition, the audio text in the current dialogue form indicates that the real-time dialogue between the user and the interactive user can be analyzed to capture the current semantic changes in a timely manner. The audio text in the dialogue form or the audio text in the non-dialogue form indicates that the pre-recorded audio can be analyzed offline to meet the semantic analysis of the user in non-real-time scenarios (such as playback).

[0019] In some preferred embodiments, when the semantic text is a text text, the semantic analysis of the semantic text includes: obtaining the text text selected by the current device user or the text text automatically recommended by the recommendation algorithm for semantic analysis to obtain the sensory feature semantics that affect the sensory state of the device user of the controlled device. It should be noted that in this embodiment, the text text can be obtained according to the user selection or the automatic recommendation mechanism for semantic analysis to meet diverse requirements. Among them, the user can manually select the text text on the device side, or the recommendation algorithm can automatically recommend the text text. When the user already has a specific text text to analyze, it can be manually selected; when the user does not have a clear target text, the algorithm can be used to screen and recommend the text text that may be related to the user's mood or preference, so as to better meet the device control requirements of the user in the text scenario.

[0020] In some preferred embodiments, when the semantic text is non-dialogue-form audio text, the semantic analysis of the semantic text includes: obtaining the non-dialogue-form audio text selected by the current device user or automatically recommended by the recommendation algorithm for semantic analysis, so as to obtain the sensory feature semantics that affect the sensory state of the device user of the controlled device. It should be noted that in actual use, the audio that the user may listen to is not necessarily in the form of dialogue (such as natural sounds, broadcasts, readings, etc.). By separately listing the processing for non-dialogue-form audio text, semantic analysis can be performed on a wider range of audio sources. This analysis of non-dialogue audio can capture potential emotional elements in the audio, so as to more comprehensively understand and influence the device user. The non-dialogue-form audio can be either actively selected by the user or automatically recommended by the recommendation algorithm. When the user is not sure what kind of audio they currently need, appropriate non-dialogue audio can be automatically recommended for semantic analysis based on their mood, preferences or historical records, and then the working mode of the controlled device can be automatically adjusted, making the device respond more quickly and accurately to the user's needs, while also reducing the burden of the user's frequent selection and operation and enhancing the interaction experience.

[0021] In some preferred embodiments, when the semantic text is dialogue-form audio text, the dialogue-form audio text is composed of the dialogue content of at least two dialogue participants in the dialogue scenario; the dialogue content includes the dialogue content of the device user using the controlled device and the dialogue content of the interactive user interacting with the device user. It should be noted that in the actual use environment, there is more than one person speaking in the audio dialogue. If only the words of the device user himself are focused on, it may not be possible to comprehensively and accurately capture his sensory state and needs. By collecting and analyzing the dialogue content of at least two dialogue participants (including the device user and the interactive user), the speech, intonation and context of both parties can be referred to at the same time, so as to more richly understand the sensory changes such as the user's psychology and emotions and the influence brought by the interaction process. At least two dialogue participants in the dialogue scenario can not only distinguish the speeches of different speakers, but also identify the content and emotions expressed by different speakers, so as to more accurately identify the source of the sensory changes of the controlled device user (whether it is due to his own emotional fluctuations or being stimulated by the words of the interactive user), thus improving the pertinence of semantic analysis and device control.

[0022] In some preferred embodiments, the semantic analysis of the semantic text includes: obtaining the conversation content of the conversation subject, and performing semantic analysis on the conversation content to obtain the semantic features of the sensory state that affect the device user of the controlled device. It should be noted that in this embodiment, obtaining the conversation content of the conversation subject is listed separately, and the language content and emotional information of each conversation subject can be identified and analyzed one by one to ensure the mastery of the complete context, and then more accurately infer the emotional state of the device user and the source of influence. In this embodiment, instead of generally identifying the emotions of all interlocutors, it focuses on identifying which semantic elements will affect the device user. Through such targeted semantic analysis, the sensory changes closely related to the user experience can be quickly identified in the conversation scenario, so as to guide the device to make corresponding automatic adjustments or feedback, and enhance the user's personalized and immersive experience.

[0023] In some preferred embodiments, when the audio text in the form of conversation is a recorded text in the form of conversation, obtaining the conversation content of the conversation subject and performing semantic analysis on the conversation content includes: obtaining the recorded text in the form of conversation selected by the current device user or the recorded text in the form of conversation automatically recommended by the recommendation algorithm, and performing semantic analysis on the obtained recorded text in the form of conversation to obtain the semantic features of the sensory state that affect the device user of the controlled device. It should be noted that compared with real-time conversation, the recorded text belongs to offline audio and can be played back later for semantic analysis. In this embodiment, through the recorded text in the form of conversation, more types of conversation scenarios can be compatible. For example, it supports both the analysis of the user's real-time call content and the analysis of the recordings manually selected by the user later or automatically recommended by the recommendation algorithm. This compatibility can significantly expand the application scope, meet the diverse scenario requirements, and enable the intelligent device to achieve emotional responses in more situations. If the user already has specific recorded content to analyze, they can select it manually; if the user does not have clear recorded content, the recommendation algorithm can automatically recommend it based on interests, emotions, historical usage habits, etc.

[0024] In some preferred embodiments, when the audio text in the dialogue form is the audio text of the current dialogue form, the dialogue content of the dialogue subject is obtained, and semantic analysis is performed on the dialogue content, including: obtaining the current dialogue content of the interactive user, and performing semantic analysis according to the current dialogue content of the interactive user to obtain the semantic of the sensory feature that affects the sensory state of the device user of the controlled device. It should be noted that different from the ex post analysis of the recorded text, the audio text of the current dialogue form emphasizes performing semantic analysis while the dialogue is in progress. By capturing and analyzing the current dialogue content of the interactive user in real time, the speech, tone, and emotional information of the interactive user can be grasped at the moment when the dialogue occurs, and the impact on the sensory state of the device user can be perceived and judged in a timely manner. This real-time analysis can automatically adjust the device working mode at the first time, significantly improving the real-time performance and immersive interaction experience of the intelligent device. Performing semantic analysis according to the current dialogue content of the interactive user can focus on identifying the key information (such as words, emotions, tone) conveyed by the interactive user, and infer the sensory impact on the device user accordingly. This process can quickly locate the cause of the user's feeling, so as to ensure that the recognition of the sensory state of the device user is more accurate and timely, and lay a foundation for the intelligent device to adjust the mode and optimize the interaction.

[0025] In some further embodiments, when the audio text in the dialogue form is the audio text of the current dialogue form, the dialogue content of the dialogue subject is obtained, and semantic analysis is performed on the dialogue content, including: obtaining the current dialogue content of the device user and the current dialogue content of the interactive user, and performing semantic analysis according to the current dialogue content of the device user and the current dialogue content of the interactive user to obtain the semantic of the sensory feature that affects the sensory state of the device user of the controlled device. It should be noted that different from only analyzing the dialogue content of the interactive user or only analyzing the dialogue content of the device user, in this embodiment, by obtaining the current dialogue content of the device user and the current dialogue content of the interactive user at the same time, a more comprehensive and rich understanding of the context of the current dialogue can be obtained. For example: What did the device user say? What was the tone? What did the interactive user say? What was the emotion? How did the languages of both sides affect each other? By collecting and analyzing the information of both parties simultaneously while the dialogue is in progress, the context and the emotional logic behind the dialogue can be understood more accurately, so as to make a more appropriate semantic judgment.

[0026] In some further embodiments, the conversation content is semantically analyzed, including: combining the conversation content of the device user and the conversation content of the interactive user into contextual content with contextual context; analyzing the contextual content to obtain the sensory feature semantics that affect the sensory state of the device user of the controlled device. It should be noted that the conversation between the device user and the interactive user is often a continuous and mutually influential process. The speech of the device user will cause the response of the interactive user, and the speech of the interactive user will in turn affect the device user. In this embodiment, the conversation content of both parties is merged into contextual content with contextual context for analysis, so that the complete conversation context can be understood, thereby more accurately identifying various factors that affect the sensory state of the device user. When the conversation content is integrated into contextual content, not only can the semantics of a single sentence or a single paragraph be identified, but also the previous and subsequent associations of the conversation can be tracked (such as whether a certain sentence responds to a previous topic, and whether a progression or opposition is formed in terms of emotions). This contextual association can distinguish short-term emotional fluctuations from continuous emotional attitudes, and judge the sensory changes of the device user (such as emotional trends or changes in psychological states), and finally automatically control the controlled device based on more accurate and rich sensory feature semantics, thereby significantly enhancing the effects of emotional interaction and personalized response.

[0027] In some further embodiments, when analyzing the context content, it includes: analyzing the emotional semantics in the context content to obtain the sensory feature semantics that affect the emotional state of the device user in the dialogue scene. It should be noted that in this embodiment, the recognition and extraction of emotional semantics are highlighted, and the direct or implicit emotional components in the dialogue scene are taken as the focus of analysis. This can accurately locate which emotions, attitudes or psychological factors can most affect the device user, and provide more valuable emotional basis for the subsequent automatic adjustment of the device. It is not enough to obtain general semantic information (such as topics, intentions, etc.) to achieve deep emotional interaction. In this embodiment, by digging deep into the emotional semantics in the dialogue, the emotional semantics are analyzed in detail, so as to extract the sensory feature semantics that affect the emotional state of the device user, and then map these results to the automatic control of different working modes of the device (such as a massager). This chain from emotional semantics to device control can greatly enhance the user's immersive experience and satisfaction, and make the response of smart devices more humane and emotional.

[0028] In some further embodiments, when analyzing the emotional semantics in the context content, it includes: parsing the semantic text that embodies the context content to obtain emotional keywords, tone keywords and context keywords in the semantic text, and analyzing the emotional semantics in the context content according to the emotional keywords, the tone keywords and the context keywords to obtain the sensory feature semantics that affect the emotional state of the device user in the dialogue scene. It should be noted that in the dialogue scene, the user's emotions are often reflected in multiple dimensions: whether the words themselves have a positive or negative tendency (emotional keywords), whether the tone contains a strong or subtle attitude (tone keywords), and how the dialogue background or context shapes this language (context keywords). In this embodiment, clues are extracted from the above multiple dimensions at the same time. In this way, the hierarchical understanding of text information can be refined, so that the emotional analysis is more in-depth and accurate in technology, and no longer stays at a single emotional label recognition. When analyzing text, if a single keyword (such as "angry" or "happy") is used to judge the user's emotions, it is often impossible to accurately distinguish the severity of the tone or the true meaning of the context. Combining emotional keywords (indicating emotional tendencies) + tone keywords (indicating expression methods and attitudes) + context keywords (indicating scenes and contextual connections) can better grasp the overall meaning and emotional direction of the conversation. This multi-angle fusion analysis method can quickly and accurately identify the semantic elements that have a key impact on the user's sensory changes, providing a more accurate basis for subsequent automatic device control (such as vibration mode, lighting adjustment, etc.), thereby enhancing the user's immersive experience and emotional identification.

[0029] In some further embodiments, the emotional semantics in the context content is analyzed by an AI emotional analysis model, and the AI ​​emotional analysis model analyzes the emotional semantics in the context content according to the emotional keywords, the tone keywords and the context keywords to obtain the sensory feature semantics that affect the emotional state of the device user in the dialogue scene. The AI ​​emotional analysis model may include any one of a convolutional neural network, a recurrent neural network, a long short-term memory network, a deep neural network and a transformer network (Transformer) for emotional analysis. It should be noted that, compared with the manually defined rule system, the AI ​​model has the ability of self-learning and generalization, and can continuously optimize the recognition accuracy in the continuously accumulated dialogue data and feedback. Through the comprehensive learning of emotional keywords, tone keywords and context keywords, the AI ​​model can analyze the user's emotional fluctuations in real time or quasi-real time and produce accurate sensory feature semantics that affect the emotional state of the device user, so that the smart device can quickly and flexibly adjust the working mode according to the user's emotional state, thereby significantly improving the user's immersive experience and intelligent interaction effect.

[0030] In some preferred embodiments, when the device user selects the text or the audio text in the form of dialogue or the audio text in the form of non-dialogue, the selection is made through the user-side interaction interface or voice commands.

[0031] Embodiment II See Figure 2 、 Figure 4 , this embodiment provides a device control device based on semantic analysis, including: A semantic analysis module for performing semantic analysis on semantic texts to obtain semantic features of the feeling state of the device user affecting the controlled device; A working mode control module for automatically controlling the working mode of the controlled device according to the semantic features of the feeling state, so that the controlled device acts on the device user in different working modes according to different semantic features of the feeling state, and changes the usage experience of the device user when using the controlled device.

[0032] In this embodiment, by performing semantic analysis on semantic texts to obtain semantic features of the feeling state of the device user affecting the controlled device, and automatically controlling the working mode of the controlled device according to the semantic features of the feeling state, the controlled device acts on the device user in different working modes according to different semantic features of the feeling state, and changes the usage experience of the device user when using the controlled device, thereby establishing the correlation degree among semantics, the controlled device, and the usage experience of the controlled device, improving the intelligent level of device control, and improving the user experience.

[0033] Embodiment III See Figure 1 , Figure 3 , Figure 4 , this embodiment provides a controlled device, and the controlled device uses any one of the device control methods based on semantic analysis in the above embodiments for control. The controlled device can be connected and communicate with the user side and the server. By performing semantic analysis on semantic texts to obtain semantic features of the feeling state of the device user affecting the controlled device, and automatically controlling the working mode of the controlled device according to the semantic features of the feeling state, the controlled device acts on the device user in different working modes according to different semantic features of the feeling state, and changes the usage experience of the device user when using the controlled device, thereby establishing the correlation degree among semantics, the controlled device, and the usage experience of the controlled device, improving the intelligent level of device control, and improving the user experience.

[0034] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements on some or all of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A device control method based on semantic analysis, characterized in that: include: Performing semantic analysis on the semantic text to obtain the sensory feature semantics that affect the sensory state of the device user of the controlled device; The working mode of the controlled device is automatically controlled according to the sensory feature semantics, so that the controlled device acts on the device user in different working modes according to the sensory feature semantics, thereby changing the user experience of the device user when using the controlled device.

2. The device control method based on semantic analysis according to claim 1, characterized in that: The semantic text is a text text or a voice text; the voice text is an audio text in a dialogue form or an audio text in a non-dialogue form; the audio text in a dialogue form is an audio text in a current dialogue form or a recorded text in a dialogue form.

3. The device control method based on semantic analysis according to claim 2, characterized in that: When the semantic text is a text text, the semantic analysis of the semantic text includes: obtaining the text text currently selected by the device user or the text text automatically recommended by the recommendation algorithm for semantic analysis to obtain the sensory feature semantics that affect the sensory state of the device user of the controlled device.

4. The device control method based on semantic analysis according to claim 2, characterized in that: When the semantic text is a non-dialog audio text, the semantic analysis of the semantic text includes: obtaining the non-dialog audio text currently selected by the device user or the non-dialog audio text automatically recommended by the recommendation algorithm, and performing semantic analysis to obtain the sensory feature semantics that affect the sensory state of the device user of the controlled device.

5. The device control method based on semantic analysis according to claim 2, characterized in that: When the semantic text is an audio text in a dialogue form, the audio text in a dialogue form is composed of dialogue contents of at least two dialogue subjects in a dialogue scene; the dialogue contents include dialogue contents of a device user using a controlled device and dialogue contents of an interactive user interacting with the device user.

6. The device control method based on semantic analysis according to claim 5, characterized in that: The semantic analysis of the semantic text includes: acquiring the conversation content of the conversation subject, and performing semantic analysis on the conversation content to obtain the sensory feature semantics that affect the sensory state of the device user of the controlled device.

7. The device control method based on semantic analysis according to claim 6, characterized in that: When the audio text in the form of a conversation is a recorded text in the form of a conversation, the conversation content of the conversation subject is obtained, and a semantic analysis is performed on the conversation content, including: obtaining the recorded text in the form of a conversation currently selected by the device user or the recorded text in the form of a conversation automatically recommended by a recommendation algorithm, and performing a semantic analysis on the obtained recorded text in the form of a conversation to obtain the sensory feature semantics that affect the sensory state of the device user of the controlled device.

8. The device control method based on semantic analysis according to claim 6, characterized in that: When the audio text in the form of a conversation is the audio text in the current conversation form, the conversation content of the conversation subject is obtained, and a semantic analysis is performed on the conversation content, including: obtaining the current conversation content of the device user and the current conversation content of the interactive user, and performing a semantic analysis based on the current conversation content of the device user and the current conversation content of the interactive user to obtain the sensory feature semantics that affect the sensory state of the device user of the controlled device.

9. The device control method based on semantic analysis according to claim 6, characterized in that: The conversation content is semantically analyzed, including: combining the conversation content of the device user and the conversation content of the interactive user into contextual content with a contextual context; analyzing the contextual content to obtain the sensory feature semantics that affect the sensory state of the device user of the controlled device.

10. The device control method based on semantic analysis according to claim 9, characterized in that: When analyzing the context content, it includes: analyzing the emotional semantics in the context content to obtain the sensory feature semantics that affect the emotional state of the device user in the dialogue scene.

11. The device control method based on semantic analysis according to claim 10, characterized in that: When analyzing the emotional semantics in the context content, it includes: parsing the semantic text that embodies the context content to obtain emotional keywords, tone keywords and context keywords in the semantic text; analyzing the emotional semantics in the context content based on the emotional keywords, the tone keywords and the context keywords to obtain the sensory feature semantics that affect the emotional state of the device user in the dialogue scene.

12. A device control device based on semantic analysis, characterized in that: include: A semantic analysis module, used for performing semantic analysis on the semantic text to obtain the sensory feature semantics that affect the sensory state of the device user of the controlled device; The working mode control module is used to automatically control the working mode of the controlled device according to the sensory feature semantics, so that the controlled device acts on the device user in different working modes according to the different sensory feature semantics, thereby changing the user experience of the device user when using the controlled device.

13. A controlled device, characterized in that: The controlled device is controlled using the device control method based on semantic analysis as described in any one of claims 1-11.

14. A server, characterized in that: include: Run the device control method based on semantic analysis as described in any one of claims 1 to 11 to control the controlled device.

Images