Control method and device of intelligent toy, electronic equipment and storage medium

Abstract

The application provides a control method of an intelligent toy, which comprises the following steps: when the intelligent toy is in a question and answer state, determining whether a target special knowledge base corresponding to a current growth value of the intelligent toy exists; if the target special knowledge base corresponding to the current growth value exists, performing first question and answer processing on question and answer data of a user based on the target special knowledge base to obtain first answer data, and controlling the intelligent toy to output the first answer data; if the target knowledge base corresponding to the current growth value does not exist, determining a target general knowledge base corresponding to age information of the user; performing second question and answer processing on the question and answer data of the user based on the target general knowledge base to obtain second answer data, and controlling the intelligent toy to output the second answer data. The matching logic of the special knowledge base and the general knowledge base can generate the corresponding knowledge of the user for output, thereby enhancing the interactive experience of the user.

Description

Technical Field

[0001] This invention relates to the field of artificial intelligence, and more particularly to a control method, device, electronic device, and storage medium for an intelligent toy. Background Technology

[0002] Current smart toys offer children storytelling abilities and simple voice command interaction, while also imparting knowledge during these interactions. However, because these smart toys rely on simple voice command interaction and are based on a fixed database, they output fixed content corresponding to the voice commands. This fixed content does not maintain children's attention for long. Therefore, existing smart toys suffer from a lack of sustained engagement, primarily relying on a fixed database and outputting fixed content corresponding to voice commands, resulting in a poor user experience. Summary of the Invention

[0003] This invention provides a control method for smart toys, aiming to solve the problems of existing smart toys having limited interaction methods, failing to continuously attract user interest, and resulting in a poor user experience. This invention generates user-specific knowledge through a matching logic between a dedicated knowledge base and a general knowledge base, enhancing the user's interactive experience and resolving the issues of limited interaction methods, failure to continuously attract user interest, and poor user experience in existing smart toys.

[0004] In a first aspect, embodiments of the present invention provide a control method for an intelligent toy, the method comprising the following steps:

[0005] When the smart toy is in a question-and-answerable state, it is determined whether there is a target-specific knowledge base corresponding to the current growth value of the smart toy. The level of the target-specific knowledge base corresponds to the current growth value of the smart toy, and different levels of target-specific knowledge bases have knowledge data corresponding to different user ages.

[0006] If a target-specific knowledge base exists corresponding to the current growth value, then based on the target-specific knowledge base, the user's question and answer data is processed in the first question and answer stage to obtain the first answer data, and the smart toy is controlled to output the first answer data;

[0007] If there is no target knowledge base corresponding to the current growth value, then a target general knowledge base corresponding to the user's age information is determined, and the different levels of the general knowledge base correspond to different ages;

[0008] Based on the target general knowledge base, the user's question and answer data is processed in a second question and answer manner to obtain second answer data, and the smart toy is controlled to output the second answer data.

[0009] Optionally, before determining whether a target-specific knowledge base corresponding to the current growth value of the smart toy exists when the smart toy is in a question-and-answer state, the method includes:

[0010] When a user's wake-up command is detected, the user's voice data is acquired;

[0011] Perform semantic recognition on the speech data to obtain the semantic recognition result of the speech data;

[0012] Based on the semantic recognition results, it is determined whether the smart toy is in a question-and-answer state.

[0013] Optionally, the first question-and-answer processing of the user's question-and-answer data based on the target-specific knowledge base to obtain first answer data includes:

[0014] Based on the target-specific knowledge base and the user's question-and-answer data, prompt words for the large language model are constructed. The prompt words are used to prompt the large language model to call the target-specific knowledge base to process the user's question-and-answer data.

[0015] The prompt words are input into the large language model for the first question-and-answer process to obtain the first answer data.

[0016] Optionally, determining the target general knowledge base corresponding to the user's age information includes:

[0017] Obtain the question and answer types from the question and answer data;

[0018] Based on the question-and-answer type, a target knowledge sub-base corresponding to the user's age information is determined in the general knowledge base. The general knowledge base includes different types of knowledge sub-bases, and the different levels of the general knowledge base correspond to knowledge data of different ages.

[0019] Optionally, the step of performing second question-and-answer processing on the user's question-and-answer data based on the target general knowledge base to obtain second answer data, and controlling the smart toy to output the second answer data, includes:

[0020] Based on the target knowledge sub-base and the user's question and answer data, prompt words for the large language model are constructed. The prompt words are used to instruct the large language model to call the target knowledge sub-base to perform question and answer processing on the question and answer data.

[0021] The prompt words are input into the large language model for the second question-and-answer process to obtain the second answer data.

[0022] Optionally, the second question-and-answer processing of the user's question-and-answer data based on the target general knowledge base to obtain second answer data includes:

[0023] Obtain the historical dialogue dataset between the user and the smart toy, the historical dialogue dataset containing at least one historical question-and-answer pair;

[0024] Based on the target knowledge sub-base, the historical dialogue dataset, and the user's question-and-answer data, a second prompt word for the large language model is constructed. The second prompt word is used to instruct the large language model to call the target knowledge sub-base to perform question-and-answer processing on the question-and-answer data.

[0025] The second prompt word is input into the large language model for the second question-and-answer process to obtain the second answer data.

[0026] Optionally, after performing second question-and-answer processing on the user's question-and-answer data based on the target general knowledge base to obtain second answer data, and controlling the smart toy to output the second answer data, the method further includes:

[0027] If there is no target general knowledge base corresponding to the user's age information, a temporary knowledge base is generated based on the online knowledge platform;

[0028] Based on the temporary knowledge base, the user's question and answer data is processed in a third question and answer manner to obtain third answer data, and the smart toy is controlled to output the third answer data.

[0029] Optionally, the step of performing third question-and-answer processing on the user's question-and-answer data to obtain third answer data, and controlling the smart toy to output the third answer data, includes:

[0030] The user's question and answer data is subjected to third-party question and answer processing to obtain the answer data to be processed;

[0031] The answer data to be processed is reviewed to determine whether the review of the answer data to be processed is successful.

[0032] If the answer data to be processed is successfully reviewed, the answer data to be processed is determined to be the third answer data, and the smart toy is controlled to output the third answer data.

[0033] Secondly, embodiments of the present invention also provide a control device for a smart toy, the control device for the smart toy comprising:

[0034] The first determining module is used to determine whether there is a target dedicated knowledge base corresponding to the current growth value of the smart toy when the smart toy is in a question-and-answer state. The level of the dedicated knowledge base corresponds to the current growth value of the smart toy, and different levels of the target dedicated knowledge base have knowledge data corresponding to different user ages.

[0035] The first control module is configured to, if a target-specific knowledge base exists corresponding to the current growth value, perform a first question-and-answer process on the user's question-and-answer data based on the target-specific knowledge base to obtain first answer data, and control the smart toy to output the first answer data;

[0036] The second determining module is used to determine a target general knowledge base corresponding to the user's age information if there is no target knowledge base corresponding to the current growth value. The different levels of the general knowledge base correspond to different ages.

[0037] The second control module is used to perform second question-and-answer processing on the user's question-and-answer data based on the target general knowledge base, obtain second answer data, and control the smart toy to output the second answer data.

[0038] Thirdly, embodiments of the present invention provide an electronic device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps in the control method for the intelligent toy provided in the embodiments of the present invention.

[0039] Fourthly, embodiments of the present invention provide a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps in the control method for the intelligent toy provided in the embodiments of the present invention.

[0040] In this embodiment of the invention, when the smart toy is in a question-and-answer state, it is determined whether a target dedicated knowledge base corresponding to the smart toy's current growth value exists. The level of the dedicated knowledge base corresponds to the smart toy's current growth value, and different levels of the target dedicated knowledge base have knowledge data corresponding to different user ages. If a target dedicated knowledge base corresponding to the current growth value exists, then based on the target dedicated knowledge base, the user's question-and-answer data undergoes a first question-and-answer process to obtain first answer data, and the smart toy is controlled to output the first answer data. If no target knowledge base corresponding to the current growth value exists, then a target general knowledge base corresponding to the user's age information is determined, and different levels of the general knowledge base correspond to different ages. Based on the target general knowledge base, the user's question-and-answer data undergoes a second question-and-answer process to obtain second answer data, and the smart toy is controlled to output the second answer data. This invention can generate user-specific knowledge through the matching logic of dedicated knowledge bases and general knowledge bases, enhancing the user's interactive experience and solving the problem that existing smart toys have a single interaction method, mainly based on a fixed database, outputting fixed content corresponding to voice commands, failing to continuously attract user interest, and resulting in a poor companionship experience. Attached Figure Description

[0041] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0042] Figure 1 This is a flowchart of a control method for an intelligent toy provided in an embodiment of the present invention;

[0043] Figure 2 This is a schematic diagram of the structure of a control device for an intelligent toy provided in an embodiment of the present invention;

[0044] Figure 3 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0045] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0046] like Figure 1 As shown, Figure 1This is a flowchart of a control method for an intelligent toy provided in an embodiment of the present invention. The control method for the intelligent toy includes the following steps:

[0047] 101. When the smart toy is in a question-and-answer state, determine whether there is a target-specific knowledge base corresponding to the smart toy's current growth value.

[0048] In this embodiment of the invention, the control method for the smart toy can be applied to a server, with a communication connection between the server and the smart toy. The smart toy includes a voice acquisition device and an image acquisition device. The voice acquisition device is used to acquire voice data, and the image acquisition device is used to acquire image data. When using the smart toy for the first time, the user can set user information, including the user's name, age, hobbies, etc. The smart toy can also be set with a wake-up word by the user, such as "Hello, Little One" or "Hello, Little Bean." The smart toy can also be set with a special title for a specific user, such as "Little Master," which the smart toy can use to address the user as "Little Master" when interacting with them.

[0049] The users mentioned above can be understood as those who wake up the smart toy.

[0050] The above question-and-answer state can be understood as the state in which the smart toy can interact with the user through question-and-answer sessions.

[0051] The levels of the aforementioned specialized knowledge base correspond to the current growth value of the smart toy, and different levels of the target specialized knowledge base have knowledge data corresponding to different user ages.

[0052] The aforementioned target-specific knowledge base can be understood as a dedicated knowledge base corresponding to the current growth value of the intelligent toy.

[0053] Furthermore, when the smart toy is in a question-and-answer state, its current growth value is obtained. Based on the correspondence between the growth value and a dedicated knowledge base, it is determined whether a target dedicated knowledge base corresponding to the smart toy's current growth value exists. The aforementioned current growth value of the smart toy can be understood as the current growth value of the smart toy through interaction with the user. The smart toy's growth value corresponds to its dedicated knowledge base, and different levels of the target dedicated knowledge base contain knowledge data corresponding to different user ages.

[0054] The aforementioned dedicated knowledge base is a pre-set knowledge base with different levels. It is generated and updated based on historical data accumulated during the user's journey, making it a user-specific knowledge base with a degree of personalization. For example, the dedicated knowledge base could be an initial level knowledge base, a first-level knowledge base, a second-level knowledge base, a third-level knowledge base, etc. The initial level knowledge base could correspond to knowledge data for a 2-year-old child and personalized data for that age group; the first-level knowledge base could correspond to knowledge data for a 2-3-year-old child and personalized data for that age group; the second-level knowledge base could correspond to knowledge data for a 3-4-year-old child and personalized data for that age group; and the third-level knowledge base could correspond to knowledge data for a 4-5-year-old child and personalized data for that age group, etc.

[0055] The growth values ​​of the aforementioned smart toys correspond to the dedicated knowledge base of the smart toys. Specifically, the initial growth value of the smart toy corresponds to the initial level of the dedicated knowledge base, the first growth value of the smart toy corresponds to the first level of the knowledge base, the second growth value of the smart toy corresponds to the second level of the dedicated knowledge base, the third growth value of the smart toy corresponds to the third level of the dedicated knowledge base, and so on. The third growth value is higher than the second growth value, the second growth value is higher than the first growth value, and the first growth value is higher than the initial growth value.

[0056] In one possible embodiment, when the current growth value of the smart toy is the initial growth value, it is determined whether there is dedicated knowledge corresponding to the current growth value based on the correspondence between growth values ​​and dedicated knowledge bases; when the current growth value of the smart toy is the first growth value, it is determined whether there is dedicated knowledge base corresponding to the current growth value based on the correspondence between growth values ​​and dedicated knowledge bases; when the current growth value of the smart toy is the second growth value, it is determined whether there is dedicated knowledge base corresponding to the current growth value based on the correspondence between growth values ​​and dedicated knowledge bases; when the current growth value of the smart toy is the third growth value, it is determined whether there is dedicated knowledge base corresponding to the current growth value based on the correspondence between growth values ​​and dedicated knowledge bases.

[0057] 102. If a target-specific knowledge base exists corresponding to the current growth value, then based on the target-specific knowledge base, perform the first question-and-answer processing on the user's question-and-answer data to obtain the first answer data, and control the smart toy to output the first answer data.

[0058] In this embodiment of the invention, when it is determined that there is a target-specific knowledge base corresponding to the current growth value, the user's question and answer data can be processed by the target-specific knowledge base to obtain the first answer data.

[0059] The question-and-answer data of the above users can be understood as data containing questions, such as "What color is an apple?" or "What is 4 plus 3?"

[0060] The aforementioned question-and-answer data is obtained by collecting users' voice data in real time through the voice acquisition device of the smart toy and performing semantic recognition on the user's voice data. The aforementioned voice data can be questions such as "What color is an apple?" or "Please tell me what 4 plus 3 equals."

[0061] The aforementioned semantic recognition can be understood as the process of semantic recognition of speech data, which can be achieved through natural language processing (NLP). The NLP technology aims to enable machines to understand, interpret, and generate human language, achieving effective communication between humans and machines, and allowing computers to perform tasks such as language translation, sentiment analysis, and text summarization.

[0062] The first question-and-answer process described above can be understood as the process of finding the most relevant answer to the question-and-answer data from a target-specific knowledge base.

[0063] The aforementioned first answer data can be understood as the most relevant answer data obtained by analyzing question and answer data.

[0064] In one possible embodiment, when the target-specific knowledge base corresponding to the current growth value is a first-level knowledge base, if the question-and-answer data is "What color is an apple?", and the target-specific knowledge base includes knowledge data about fruits, the answer data "Apples are red" can be found in the first-level knowledge base for the question-and-answer data "What color is an apple?", and the smart toy can be controlled to output the answer data "Apples are red". When the target-specific knowledge base corresponding to the current growth value is a third-level knowledge base, if the question-and-answer data is "What does 4 plus 3 equal?", and the target-specific knowledge base includes knowledge data about addition and subtraction within 100, the answer data "4 plus 3 equals 7" can be found in the third-level knowledge base for the question-and-answer data "What does 4 plus 3 equal?", and the smart toy can be controlled to output the answer data "4 plus 3 equals 7".

[0065] The above output can be understood as a pre-set output format by the system, which can be either voice output or action output. Voice output means the smart toy converts answer data into audible voice information. Action output means the smart toy performs corresponding physical actions based on the answer data, such as moving its eyes, rotating its arms, or moving other parts. Through action output, users can obtain answers in a direct or interactive way.

[0066] 103. If there is no target knowledge base corresponding to the current growth value, then determine the target general knowledge base corresponding to the user's age information.

[0067] In this embodiment of the invention, the general knowledge base does not include the user's personalized data, and different levels of the general knowledge base correspond to different ages. It should be noted that if there is no target knowledge base corresponding to the current growth value, it indicates that the user is using the smart toy for the first time, or that the user is a non-bound user, meaning that the user does not have a dedicated knowledge base.

[0068] Furthermore, if it is determined that there is no target knowledge base corresponding to the current growth value, the user's age information can be obtained, and the target general knowledge base corresponding to the user's age information can be determined by matching the user's age information in the general knowledge database.

[0069] The age information of the users mentioned above is the age information set by the users when using the smart toys, such as 5, 7, 8, etc.

[0070] The aforementioned target general knowledge base can be understood as a general knowledge base corresponding to user age information.

[0071] The aforementioned general knowledge base is a pre-set general knowledge base with different levels. For example, it can be a first-level knowledge base, a second-level knowledge base, a third-level knowledge base, a fourth-level knowledge base, a fifth-level knowledge base, etc. The first-level knowledge base corresponds to knowledge data for children aged 2-3; the second-level knowledge base corresponds to knowledge data for children aged 4-5; the third-level knowledge base corresponds to knowledge data for children aged 6; the fourth-level knowledge base corresponds to knowledge data for children aged 7; and the fifth-level knowledge base corresponds to knowledge data for children aged 8, etc.

[0072] In one possible embodiment, when no target knowledge base corresponding to the current growth value exists, and the user's age information is 4 years old, a match is made in the general knowledge base based on the user's age information 4 to determine the target general knowledge base corresponding to the user's age information 4 as the second-level knowledge base; when no target knowledge base corresponding to the current growth value exists, and the user's age information is 6 years old, a match is made in the general knowledge base based on the user's age information 6 to determine the target general knowledge base corresponding to the user's age information 6 as the third-level knowledge base; when no target knowledge base corresponding to the current growth value exists, and the user's age information is 8 years old, a match is made in the general knowledge base based on the user's age information 8 to determine the target general knowledge base corresponding to the user's age information 8 as the fifth-level knowledge base.

[0073] In another possible embodiment, when using the smart toy for the first time, the user can set their age to n, where n can be 4, 5, 6, etc. It is understood that after the user sets their age, the smart toy can match the age to a corresponding level of a general knowledge base. The smart toy can then interact with the user using the knowledge data from this level's knowledge base, effectively improving the companionship experience. When the user is linked to the smart toy, a dedicated knowledge base corresponding to that user's age group will be generated based on the general knowledge base, used to store the user's personalized data.

[0074] 104. Based on the target general knowledge base, perform second question-and-answer processing on the user's question-and-answer data to obtain second answer data, and control the smart toy to output the second answer data.

[0075] In this embodiment of the invention, the user's question and answer data can be understood as data containing questions, such as "What color is an apple?" or "What is 4 times 3?"

[0076] The aforementioned question-and-answer data is obtained by collecting users' voice data in real time through the voice acquisition device of the smart toy and performing semantic recognition on the user's voice data. The aforementioned voice data can be questions such as "How does a snail walk?" or "Please tell me what 4 times 3 equals?"

[0077] The aforementioned semantic recognition can be understood as the process of semantic recognition of speech data, which can be achieved through natural language processing (NLP). The NLP technology aims to enable machines to understand, interpret, and generate human language, achieving effective communication between humans and machines, and allowing computers to perform tasks such as language translation, sentiment analysis, and text summarization.

[0078] The second question-and-answer process described above can be understood as the process of finding the most relevant answer to the question-and-answer data from the target general knowledge base.

[0079] The second answer data mentioned above can be understood as the most relevant answer data obtained by analyzing the question and answer data.

[0080] The above output can be understood as a pre-set output format by the system, which can be either voice output or action output. Voice output means the smart toy converts answer data into audible voice information. Action output means the smart toy performs corresponding physical actions based on the answer data, such as moving its eyes, rotating its arms, or moving other parts. Through action output, users can obtain answers in a direct or interactive way.

[0081] In one possible embodiment, when the target general knowledge base is a second-level knowledge base, if the question-and-answer data is "What color is an apple?", and the second-level knowledge base includes knowledge data about fruits, the answer data "Apples are red" can be found in the second-level knowledge base for the question-and-answer data "What color is an apple?", and the smart toy can be controlled to output the answer data "Apples are red". When the target general knowledge base is a fifth-level knowledge base, if the question-and-answer data is "What is 4 multiplied by 3?", and the target specific knowledge base includes knowledge data about addition, subtraction, multiplication, and division within 100, the answer data "4 multiplied by 3 equals 12" can be found in the fifth-level knowledge base for the question-and-answer data "What is 4 multiplied by 3?", and the smart toy can be controlled to output the answer data "4 multiplied by 3 equals 12".

[0082] In this embodiment of the invention, when the smart toy is in a question-and-answer state, it is determined whether a target dedicated knowledge base corresponding to the smart toy's current growth value exists. If a target dedicated knowledge base corresponding to the current growth value exists, then based on the target dedicated knowledge base, the user's question-and-answer data undergoes a first question-and-answer process to obtain first answer data, and the smart toy is controlled to output the first answer data. If no target knowledge base corresponding to the current growth value exists, then a target general knowledge base corresponding to the user's age information is determined, with different levels of general knowledge bases corresponding to different ages. Based on the target general knowledge base, the user's question-and-answer data undergoes a second question-and-answer process to obtain second answer data, and the smart toy is controlled to output the second answer data. This invention can generate knowledge corresponding to the user through the matching logic of dedicated knowledge bases and general knowledge bases, enhancing the user's interactive experience and solving the problem that existing smart toys have a single interaction method, mainly based on a fixed database, outputting fixed content corresponding to voice commands, failing to continuously attract user interest, and resulting in a poor companionship experience.

[0083] It is understood that in the specific implementation of this application, data such as voice data, image data, knowledge data, and user data are involved. When the embodiments in this application are applied to specific products or technologies, user permission or consent is required. Furthermore, the collection, use, and processing of related data, as well as the training, deployment, and invocation of algorithm models, must comply with the relevant laws, regulations, and standards of the relevant countries and regions.

[0084] Optionally, before determining whether a target-specific knowledge base corresponding to the smart toy's current growth value exists when the smart toy is in a question-and-answer state, the user's voice data can be acquired when the user's wake-up command is detected; semantic recognition can be performed on the voice data to obtain the semantic recognition result of the voice data; and based on the semantic recognition result, it can be determined whether the smart toy is in a question-and-answer state.

[0085] In this embodiment of the invention, when the smart toy is used for the first time, the user can set a wake-up word, such as "Hello, Little One" or "Hello, Little Bean." The wake-up command can be understood as an instruction for the user to activate the smart toy in a specific way. The user can wake up the smart toy using the set wake-up word, or through touch wake-up operations or other wake-up methods.

[0086] The users mentioned above can be understood as those who wake up the smart toy.

[0087] The user's voice data was collected through the voice acquisition device of the smart toy. This voice data could include phrases such as "What color is an apple?", "What is 4 plus 3?", or "Can I give you something to eat?"

[0088] The aforementioned semantic recognition can be understood as the process of semantic recognition of speech data. Natural Language Processing (NLP) technology can be used to perform semantic recognition processing on speech data. NLP technology enables machines to understand, interpret, and generate human language, achieving effective communication between humans and machines, and allowing computers to perform tasks such as language translation, sentiment analysis, and text summarization.

[0089] The semantic recognition results described above are the understanding obtained by analyzing speech data using natural language processing technology. These results include the user's intent expressed in their speech. For example, if a user says "What color is an apple?", the semantic recognition result indicates an intent to interact with the smart toy through a question-and-answer session; if a user says "Can I give you something to eat?", the semantic recognition result indicates an intent to feed the smart toy, and so on. These user intents can be understood as the purpose the user wants to express or the action they want to perform. User intents can be feeding intents, question-and-answer intents, etc. The feeding intent described above can be understood as the user's intent to interact with the smart toy by feeding it. The question-and-answer intent described above can be understood as the user's intent to interact with the smart toy by asking and answering questions.

[0090] Furthermore, based on the user's intent, it can be determined whether the smart toy is in a question-and-answer state.

[0091] The above question-and-answer state can be understood as the state in which the smart toy can interact with the user through question-and-answer sessions.

[0092] In one possible embodiment, when the wake word of the smart toy is "Hello, Little Bean," when the user says to the smart toy, "Hello, Little Bean, what color is an apple?", the smart toy is awakened by the wake word "Hello, Little Bean." Then, semantic recognition is performed on the user's voice data "what color is an apple?" to obtain the semantic recognition result of "intent to engage in question-and-answer interaction with the smart toy." The user's intent is determined to be a question-and-answer intent, and based on the user's question-and-answer intent, it is determined that the smart toy is in a question-and-answer state.

[0093] In another possible embodiment, when the wake word of the smart toy is "Hello, little one", when the user says to the smart toy, "Hello, little one, I want to feed you something", the smart toy is awakened by the wake word "Hello, little one". The smart toy collects the user's voice data, performs semantic recognition on the user's voice data "I want to feed you something", and obtains the semantic recognition result of "the intention to interact with the smart toy for feeding". It determines that the user's intention is a feeding intention. Based on the user's feeding intention, it is determined that the smart toy is not in a question-and-answer state.

[0094] Optionally, in the step of performing the first question-and-answer processing on the user's question-and-answer data based on the target-specific knowledge base to obtain the first answer data, prompt words for a large language model can be constructed based on the target-specific knowledge base and the user's question-and-answer data; the prompt words are then input into the large language model for the first question-and-answer processing to obtain the first answer data.

[0095] In this embodiment of the invention, the aforementioned prompt words are used to prompt the large language model to call the target dedicated knowledge base to perform question-and-answer processing on the user's question-and-answer data. The aforementioned large language model can be a large language model (LLM) built based on deep learning or machine learning, such as GPT, BERT, LLaMA, etc.

[0096] The aforementioned target-specific knowledge base refers to the dedicated knowledge base corresponding to the current growth value of the smart toy. It's important to understand that the smart toy's growth value corresponds to its dedicated knowledge base level.

[0097] The question-and-answer data mentioned above refers to data containing questions, such as "What color is an apple?" or "What is 4 plus 3?"

[0098] The first question-and-answer process described above can be understood as the process of finding the most relevant answer to the question-and-answer data from a target-specific knowledge base.

[0099] The aforementioned first answer data can be understood as the most relevant answer data obtained by analyzing question and answer data.

[0100] In one possible implementation, if the question-and-answer data is "What color is an apple?", and the target-specific knowledge base includes information about fruits, a prompt word for the large language model is constructed using the question-and-answer data and the target-specific knowledge base. The prompt word could be "Please use information from the target-specific knowledge base to answer 'What color is an apple?'". Then, the prompt word "Please use information from the target-specific knowledge base to answer 'What color is an apple?'" is input into the large language model. Based on "Please use information from the target-specific knowledge base to answer 'What color is an apple?'", the large language model retrieves the answer data "The apple is red" from the target-specific knowledge base.

[0101] In another possible embodiment, if the question-and-answer data is "What does 4 plus 3 equal?", and the target-specific knowledge base includes information about addition and subtraction within 100, a prompt word for the large language model is constructed using the question-and-answer data and the target-specific knowledge base. The prompt word could be "Please use information from the target-specific knowledge base to answer 'What does 4 plus 3 equal?'". Then, the prompt word "Please use information from the target-specific knowledge base to answer 'What does 4 plus 3 equal?'" is input into the large language model. Based on "Please use information from the target-specific knowledge base to answer 'What does 4 plus 3 equal?'", the large language model will obtain the answer data '4 plus 3 equals 7' from the target-specific knowledge base.

[0102] Optionally, in the step of determining the target general knowledge base corresponding to the user's age information, the question and answer types in the question and answer data can be obtained; based on the question and answer types, the target knowledge sub-base corresponding to the user's age information can be determined in the general knowledge base.

[0103] In this embodiment of the invention, the aforementioned general knowledge base includes different types of knowledge sub-bases, with different levels of the general knowledge base corresponding to knowledge data for different age groups. These knowledge sub-bases include habit-forming knowledge sub-bases, emotion-guiding knowledge sub-bases, and daily communication knowledge sub-bases, etc.

[0104] The above question and answer types can be understood as the types of questions raised by users, such as habit formation, emotion guidance, and daily communication.

[0105] The age information of the users mentioned above is the age information set by the users when using the smart toys, such as 1, 2, 3, 4, 5, etc.

[0106] The aforementioned general knowledge base is a pre-set general knowledge base with different levels. For example, it can be a first-level knowledge base, a second-level knowledge base, a third-level knowledge base, a fourth-level knowledge base, a fifth-level knowledge base, etc. The first-level knowledge base corresponds to knowledge data for children aged 2-3 years old, and includes sub-bases on habit formation, emotion guidance, and daily communication for children aged 2-3 years old. The second-level knowledge base corresponds to knowledge data for children aged 4-5 years old, and includes sub-bases on habit formation, emotion guidance, and daily communication for children aged 4-5 years old. The third-level knowledge base corresponds to knowledge data for children aged 6 years old, and includes... The knowledge base includes sub-bases for six-year-old children on habit formation, emotional guidance, and daily communication. The fourth-level knowledge base corresponds to knowledge data for seven-year-old children and includes sub-bases for the same categories. The fifth-level knowledge base corresponds to knowledge data for eight-year-old children and includes sub-bases for the same categories.

[0107] The aforementioned target knowledge sub-base can be understood as a knowledge sub-base within the target general knowledge base that corresponds to the user's question and answer type.

[0108] In one possible embodiment, when the question-and-answer type is habit formation, and the user's age is 4, a target general knowledge base corresponding to the user's age 4 is determined as a second-level knowledge base, and a target knowledge sub-base of the corresponding type is determined as a habit formation type knowledge sub-base within the second-level knowledge base; when the question-and-answer type is emotion guidance, and the user's age is 5, a target general knowledge base corresponding to the user's age 5 is determined as a third-level knowledge base, and a target knowledge sub-base of the corresponding type is determined as an emotion guidance type knowledge sub-base within the third-level knowledge base; when the question-and-answer type is daily communication, and the user's age is 7, a target general knowledge base corresponding to the user's age 7 is determined as a fourth-level knowledge base, and a target knowledge sub-base of the corresponding type is determined as a daily communication type knowledge sub-base within the fourth-level knowledge base.

[0109] Optionally, in the steps of performing second question-and-answer processing on the user's question-and-answer data based on the target general knowledge base to obtain second answer data, and controlling the smart toy to output the second answer data, the first prompt words of the large language model can be constructed based on the target knowledge sub-base and the user's question-and-answer data; the first prompt words are input into the large language model for second question-and-answer processing to obtain the second answer data.

[0110] In this embodiment of the invention, the first prompt word is used to instruct the large language model to call the target knowledge sub-base to perform question-and-answer processing on the question-and-answer data.

[0111] The aforementioned large language model can be a large language model built based on deep learning or machine learning, such as GPT, BERT, LLaMA, etc.

[0112] The aforementioned target knowledge sub-base is a knowledge sub-base within the target general knowledge base that corresponds to the user's question and answer type.

[0113] The question-and-answer data mentioned above refers to data containing questions, such as "What time should I get up tomorrow morning?", "It's raining, and I'm a little unhappy", and "What is 4 plus 3?".

[0114] The second question-and-answer processing described above can be understood as the process of finding the most relevant answer to the question-and-answer data from the target knowledge sub-base.

[0115] The second answer data mentioned above can be understood as the most relevant answer data obtained by analyzing the question and answer data.

[0116] In one possible implementation, if the question-and-answer data is "What time should I get up tomorrow morning?", and the target knowledge sub-base includes information about daily routines, a first prompt word is constructed using the question-and-answer data and the target knowledge sub-base to create a large language model. This first prompt word could be "Please use information from the target knowledge sub-base to answer 'What time should I get up tomorrow morning?'". Then, the first prompt word "Please use information from the target knowledge sub-base to answer 'What time should I get up tomorrow morning?'" is input into the large language model. Based on this prompt, the large language model will retrieve the answer data "I will get up at 7:30 tomorrow morning" from the target knowledge sub-base.

[0117] In another possible implementation, if the question-and-answer data is "It's raining, and I'm a little unhappy," and the target knowledge sub-base includes information about emotion guidance, a first prompt word is constructed from the question-and-answer data and the target knowledge sub-base to build a large language model. The first prompt word could be "Please use information from the target knowledge sub-base to answer 'It's raining, and I'm a little unhappy'." Then, the first prompt word "Please use information from the target knowledge sub-base to answer 'It's raining, and I'm a little unhappy'" is input into the large language model. Based on "Please use information from the target knowledge sub-base to answer 'It's raining, and I'm a little unhappy'," the large language model will retrieve the answer data "It's okay, rainy days also have their beauty; we can try listening to the rain and feeling its tranquility and rhythm" from the target knowledge sub-base.

[0118] In another possible embodiment, if the question-and-answer data is "What does 4 plus 3 equal?", and the target knowledge sub-base includes information about addition and subtraction within 100, a first prompt word is constructed for the large language model using the question-and-answer data and the target knowledge sub-base. The first prompt word could be "Please use information from the target knowledge sub-base to answer 'What does 4 plus 3 equal?'". Then, the first prompt word "Please use information from the target knowledge sub-base to answer 'What does 4 plus 3 equal?'" is input into the large language model. Based on "Please use information from the target knowledge sub-base to answer 'What does 4 plus 3 equal?'", the large language model will obtain the answer data '4 plus 3 equals 7' from the target knowledge sub-base.

[0119] Optionally, in the step of performing second question-and-answer processing on the user's question-and-answer data based on the target general knowledge base to obtain second answer data, a historical dialogue dataset between the user and the smart toy can be obtained, which contains at least one historical question-and-answer pair; based on the target knowledge sub-base, the historical dialogue dataset, and the user's question-and-answer data, a second prompt word for the large language model can be constructed, which is used to instruct the large language model to call the target knowledge sub-base to perform question-and-answer processing on the question-and-answer data; the second prompt word is input into the large language model for second question-and-answer processing to obtain the second answer data.

[0120] In this embodiment of the invention, the aforementioned historical dialogue dataset contains at least one historical question-and-answer pair. The aforementioned historical dialogue dataset can be understood as all dialogue data records generated during a user's interaction with the smart toy over a past period.

[0121] It is possible to extract relevant information related to the current user's question and answer data from historical dialogue datasets. This relevant information includes historical question and answer data and corresponding answer data, such as the user's question and the smart toy's answer.

[0122] The aforementioned target knowledge sub-base can be understood as a knowledge sub-base within the target general knowledge base that corresponds to the user's question and answer type.

[0123] The second prompt word mentioned above is used to instruct the large language model to call the target knowledge sub-base to perform question-and-answer processing on the question-and-answer data.

[0124] It should be noted that by acquiring all dialogue data generated when users interact with smart toys over a period of time, analyzing historical dialogue data, identifying related information with the current user's question and answer data, and constructing prompt words for the large language model based on knowledge in the target knowledge sub-base, combined with the related information extracted from the historical dialogue dataset and the user's question and answer data.

[0125] In one possible implementation, when the current user's question-and-answer data is mathematical, the corresponding target knowledge sub-base is also a mathematical knowledge sub-base. This sub-base can retrieve the user's dialogues with the smart toy regarding mathematical questions over a past period, including all dialogue records related to mathematical questions. Then, relevant information related to the current user's mathematical question-and-answer data can be extracted from all these dialogue records. This relevant information includes the user's question data, the smart toy's answer data, and the user's feedback. For example, if the user's question is "What is 2+2?", the smart toy's answer is "4", and the user's feedback is "Yes", then this set of information can be used as relevant information. Finally, the mathematical knowledge sub-base, relevant information, and mathematical question-and-answer data are processed together to construct prompts for the large language model. These prompts could be phrases like "Please use information from the target knowledge sub-base to answer 'What is 2+2?'". This processing can be understood as processing the relationships between questions to obtain the answer.

[0126] After obtaining the second prompt word, it can be input into the large language model for second question-and-answer processing to obtain the second answer data.

[0127] Optionally, after the steps of performing second question-and-answer processing on the user's question-and-answer data based on the target general knowledge base to obtain second answer data, and controlling the smart toy to output the second answer data, if there is no target general knowledge base corresponding to the user's age information, a temporary knowledge base is generated based on the online knowledge platform; based on the temporary knowledge base, the user's question-and-answer data is processed thirdly to obtain third answer data, and the smart toy is controlled to output the third answer data.

[0128] In this embodiment of the invention, the aforementioned online knowledge platform can be understood as an online knowledge platform implemented through network technology, allowing users to access knowledge-based question-and-answer services via the internet. This online knowledge platform includes internet search engines, professional databases, cloud knowledge bases, etc.

[0129] The aforementioned temporary knowledge base can be understood as a temporarily constructed knowledge repository.

[0130] The third question-and-answer process described above can be understood as the process of finding the most relevant answer to the question-and-answer data from a temporary knowledge base.

[0131] The third answer data mentioned above can be understood as the most relevant answer data obtained by analyzing question and answer data.

[0132] The above output can be understood as a pre-set output format by the system, which can be either voice output or action output. Voice output means the smart toy converts answer data into audible voice information. Action output means the smart toy performs corresponding physical actions based on the answer data, such as moving its eyes, rotating its arms, or moving other parts. Through action output, users can obtain answers in a direct or interactive way.

[0133] It should be noted that prompt words for the large language model can be constructed based on a temporary knowledge base and question-and-answer data. These prompt words are then input into the large language model for question-and-answer processing to obtain the answer data. The aforementioned prompt words instruct the large language model to call the target knowledge base to process the question-and-answer data. The large language model can be a large language model (LLM) built based on deep learning or machine learning, such as GPT, BERT, or LLaMA.

[0134] Optionally, in the step of performing third question-and-answer processing on the user's question-and-answer data to obtain third answer data, and controlling the smart toy to output the third answer data, the user's question-and-answer data can be processed to obtain answer data to be processed; the answer data to be processed can be reviewed to determine whether the review of the answer data to be processed is successful; if the review of the answer data to be processed is successful, the answer data to be processed is determined to be the third answer data, and the smart toy is controlled to output the third answer data.

[0135] In this embodiment of the invention, the third question-and-answer process described above can be understood as a question-and-answer process that finds the most relevant answer to the question-and-answer data from a temporary knowledge base. The temporary knowledge base is a knowledge base temporarily constructed by an online knowledge platform.

[0136] The above-mentioned answer data to be processed can be understood as answer data that needs to be reviewed.

[0137] The above review can be understood as a process of verifying the accuracy and compliance of the answer data to be processed.

[0138] The third answer data mentioned above is the most relevant answer data obtained after analyzing the question and answer data and passing the review.

[0139] The outputs described above are pre-set by the system and can be either voice or motion outputs. Voice output means the smart toy converts the answer data into audible speech. Motion output means the smart toy performs corresponding physical actions based on the answer data, such as moving its eyes, rotating its arms, or moving other parts. Through motion output, users can obtain answers in a direct or interactive way.

[0140] It should be noted that when the growth value of the smart toy reaches the preset growth value for level upgrade, the smart toy's level can be upgraded. The growth value corresponds to the smart toy's level. During each level upgrade, the smart toy can announce via voice, "Little master, I've grown a little more, and now I've learned XX skill." For example, when the smart toy's level is 0 and it upgrades to 1, it announces, "Little master, I've grown a little more, and now I can eat fruit. Please talk to me so I can learn more about you." When the smart toy's level is 1 and it upgrades to 2, it announces, "Little master, I've grown a little more, and now I can eat vegetables. I've already met my family members," and so on. This invention allows the growth value to increase with user interaction with the smart toy, enhancing its fun factor.

[0141] like Figure 2 As shown, an embodiment of the present invention provides a control device for a smart toy, the control device for the smart toy comprising:

[0142] The first determining module 201 is used to determine whether there is a target-specific knowledge base corresponding to the current growth value of the smart toy when the smart toy is in a question-and-answerable state.

[0143] The first control module 202 is used to, if there is a target-specific knowledge base corresponding to the current growth value, perform a first question-and-answer processing on the user's question-and-answer data based on the target-specific knowledge base to obtain first answer data, and control the smart toy to output the first answer data;

[0144] The second determining module 203 is used to determine a target general knowledge base corresponding to the user's age information if there is no target knowledge base corresponding to the current growth value. The different levels of the general knowledge base correspond to different ages.

[0145] The second control module 204 is used to perform second question-and-answer processing on the user's question-and-answer data based on the target general knowledge base, obtain second answer data, and control the smart toy to output the second answer data.

[0146] Optionally, the device further includes:

[0147] The acquisition module is used to acquire the user's voice data when the user's wake-up command is detected;

[0148] The semantic recognition module is used to perform semantic recognition on the speech data to obtain the semantic recognition result of the speech data;

[0149] The third determining module is used to determine whether the smart toy is in a question-and-answer state based on the semantic recognition result.

[0150] Optionally, the first control module 202 includes:

[0151] The first construction submodule is used to construct prompt words for the large language model based on the target dedicated knowledge base and the user's question and answer data. The prompt words are used to prompt the large language model to call the target dedicated knowledge base to perform question and answer processing on the user's question and answer data.

[0152] The first question-and-answer processing submodule is used to input the prompt words into the large language model for the first question-and-answer processing to obtain the first answer data.

[0153] Optionally, the second determining module 203 includes:

[0154] The acquisition submodule is used to acquire the question and answer types in the question and answer data;

[0155] The first determining submodule is used to determine, based on the question-and-answer type, the target knowledge sub-base corresponding to the user's age information in the general knowledge base. The general knowledge base includes different types of knowledge sub-bases, and the different levels of the general knowledge base correspond to knowledge data of different ages.

[0156] Optionally, the control module 204 includes:

[0157] The second construction submodule is used to construct prompt words for the large language model based on the target knowledge sub-base and the user's question and answer data. The prompt words are used to instruct the large language model to call the target knowledge sub-base to perform question and answer processing on the question and answer data.

[0158] The second question-and-answer processing submodule is used to input the prompt words into the large language model for second question-and-answer processing to obtain second answer data.

[0159] Optionally, the second construction submodule includes:

[0160] The acquisition unit is used to acquire the historical dialogue dataset between the user and the smart toy, the historical dialogue dataset containing at least one historical question-and-answer pair;

[0161] An extraction unit is used to extract association information with the user's question-and-answer data from the historical dialogue dataset, the association information including historical question-and-answer data and corresponding answer data;

[0162] The construction unit is used to construct prompt words for the large language model based on the target knowledge sub-base, the associated information, and the user's question and answer data. The prompt words are used to instruct the large language model to call the target knowledge sub-base to perform question and answer processing on the question and answer data.

[0163] Optionally, the device further includes:

[0164] The generation module is used to generate a temporary knowledge base based on the online knowledge platform if there is no target general knowledge base corresponding to the user's age information.

[0165] The third control module is used to perform third question-and-answer processing on the user's question-and-answer data based on the temporary knowledge base, obtain third answer data, and control the smart toy to output the third answer data.

[0166] Optionally, the third control module includes:

[0167] The third question-and-answer processing submodule is used to perform third question-and-answer processing on the user's question-and-answer data to obtain the answer data to be processed.

[0168] The second determining submodule is used to review the answer data to be processed and determine whether the answer data to be processed has been successfully reviewed.

[0169] The control submodule is used to determine that the answer data to be processed is the third answer data if the review of the answer data to be processed is successful, and to control the smart toy to output the third answer data.

[0170] like Figure 3 As shown, this embodiment of the invention also provides an electronic device, including a processor, which can execute any of the above-described control methods for intelligent toys.

[0171] Specifically, it includes a processor 301 and a memory 302, as well as a computer program stored in the memory 302 and capable of running on the processor 301, which executes the control method for the intelligent toy, wherein:

[0172] The processor 301 executes the calculator program for the control method of the intelligent toy stored in the memory 302, and performs the following steps:

[0173] When the smart toy is in a question-and-answerable state, determine whether there exists a target-specific knowledge base II corresponding to the smart toy's current growth value.

[0174] If a target-specific knowledge base exists corresponding to the current growth value, then based on the target-specific knowledge base, the user's question and answer data is processed in the first question and answer stage to obtain the first answer data, and the smart toy is controlled to output the first answer data;

[0175] If there is no target knowledge base corresponding to the current growth value, then a target general knowledge base corresponding to the user's age information is determined, and the different levels of the general knowledge base correspond to different ages;

[0176] Based on the target general knowledge base, the user's question and answer data is processed in a second question and answer manner to obtain second answer data, and the smart toy is controlled to output the second answer data.

[0177] Optionally, before determining whether a target-specific knowledge base corresponding to the current growth value of the smart toy exists when the smart toy is in a question-and-answer state, the method executed by the processor 301 further includes:

[0178] When a user's wake-up command is detected, the user's voice data is acquired;

[0179] Perform semantic recognition on the speech data to obtain the semantic recognition result of the speech data;

[0180] Based on the semantic recognition results, it is determined whether the smart toy is in a question-and-answer state.

[0181] Optionally, the processor 301 executes the first question-and-answer processing based on the target-specific knowledge base to obtain first answer data, including:

[0182] Based on the target-specific knowledge base and the user's question-and-answer data, prompt words for the large language model are constructed. The prompt words are used to prompt the large language model to call the target-specific knowledge base to process the user's question-and-answer data.

[0183] The prompt words are input into the large language model for the first question-and-answer process to obtain the first answer data.

[0184] Optionally, the process executed by processor 301 to determine the target general knowledge base corresponding to the user's age information includes:

[0185] Obtain the question and answer types from the question and answer data;

[0186] Based on the question-and-answer type, a target knowledge sub-base corresponding to the user's age information is determined in the general knowledge base. The general knowledge base includes different types of knowledge sub-bases, and the different levels of the general knowledge base correspond to knowledge data of different ages.

[0187] Optionally, the processor 301 executes the second question-and-answer processing based on the target general knowledge base, performs second question-and-answer data on the user's question-and-answer data to obtain second answer data, and controls the smart toy to output the second answer data, including:

[0188] Based on the target knowledge sub-base and the user's question and answer data, prompt words for the large language model are constructed. The prompt words are used to instruct the large language model to call the target knowledge sub-base to perform question and answer processing on the question and answer data.

[0189] The prompt words are input into the large language model for the second question-and-answer process to obtain the second answer data.

[0190] Optionally, the prompt words for constructing a large language model based on the target knowledge sub-base and the user's question-and-answer data, executed by processor 301, include:

[0191] Obtain the historical dialogue dataset between the user and the smart toy, the historical dialogue dataset containing at least one historical question-and-answer pair;

[0192] Extract association information with the user's question-and-answer data from the historical dialogue dataset, the association information including historical question-and-answer data and corresponding answer data;

[0193] Based on the target knowledge sub-base, the associated information, and the user's question-and-answer data, prompt words for the large language model are constructed. The prompt words are used to instruct the large language model to call the target knowledge sub-base to perform question-and-answer processing on the question-and-answer data.

[0194] Optionally, after performing second question-and-answer processing on the user's question-and-answer data based on the target general knowledge base to obtain second answer data, and controlling the smart toy to output the second answer data, the method executed by the processor 301 further includes:

[0195] If there is no target general knowledge base corresponding to the user's age information, a temporary knowledge base is generated based on the online knowledge platform;

[0196] Based on the temporary knowledge base, the user's question and answer data is processed in a third question and answer manner to obtain third answer data, and the smart toy is controlled to output the third answer data.

[0197] Optionally, the processor 301 performs third question-and-answer processing on the user's question-and-answer data to obtain third answer data, and controls the smart toy to output the third answer data, including:

[0198] The user's question and answer data is subjected to third-party question and answer processing to obtain the answer data to be processed;

[0199] The answer data to be processed is reviewed to determine whether the review of the answer data to be processed is successful.

[0200] If the answer data to be processed is successfully reviewed, the answer data to be processed is determined to be the third answer data, and the smart toy is controlled to output the third answer data.

[0201] This invention also provides a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, it implements the various processes of the intelligent toy control method provided in this invention and achieves the same technical effect. To avoid repetition, it will not be described again here.

[0202] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. The storage medium can be a magnetic disk, optical disk, read-only memory (ROM), or random access memory (RAM), etc.

[0203] The above description discloses only preferred embodiments of the present invention and should not be construed as limiting the scope of the present invention. Therefore, equivalent variations made in accordance with the claims of the present invention are still within the scope of the present invention.

Claims

1. A control method for an intelligent toy, characterized in that, The method includes the following steps: When the smart toy is in a question-and-answerable state, determine whether there is a target-specific knowledge base corresponding to the smart toy's current growth value; If a target-specific knowledge base exists corresponding to the current growth value, then based on the target-specific knowledge base, the user's question and answer data is processed in the first question and answer stage to obtain the first answer data, and the smart toy is controlled to output the first answer data; If there is no target knowledge base corresponding to the current growth value, then a target general knowledge base corresponding to the user's age information is determined, and the different levels of the general knowledge base correspond to different ages; Based on the target general knowledge base, the user's question and answer data is processed in a second question and answer manner to obtain second answer data, and the smart toy is controlled to output the second answer data.

2. The method as described in claim 1, characterized in that, Before determining whether a target-specific knowledge base corresponding to the current growth value of the smart toy exists when the smart toy is in a question-and-answer state, the method includes: When a user's wake-up command is detected, the user's voice data is acquired; Perform semantic recognition on the speech data to obtain the semantic recognition result of the speech data; Based on the semantic recognition results, it is determined whether the smart toy is in a question-and-answer state.

3. The control method for the intelligent toy as described in claim 2, characterized in that, The first question-and-answer processing, based on the target-specific knowledge base, on the user's question-and-answer data to obtain first answer data, includes: Based on the target-specific knowledge base and the user's question-and-answer data, prompt words for the large language model are constructed. The prompt words are used to prompt the large language model to call the target-specific knowledge base to process the user's question-and-answer data. The prompt words are input into the large language model for the first question-and-answer process to obtain the first answer data.

4. The control method for the intelligent toy as described in claim 3, characterized in that, The step of determining the target general knowledge base corresponding to the user's age information includes: Obtain the question and answer types from the question and answer data; Based on the question-and-answer type, a target knowledge sub-base corresponding to the user's age information is determined in the general knowledge base. The general knowledge base includes different types of knowledge sub-bases, and the different levels of the general knowledge base correspond to knowledge data of different ages.

5. The control method for the intelligent toy as described in claim 4, characterized in that, The second question-and-answer processing, based on the target general knowledge base, on the user's question-and-answer data to obtain second answer data includes: Based on the target knowledge sub-base and the user's question and answer data, a first prompt word for the large language model is constructed. The first prompt word is used to instruct the large language model to call the target knowledge sub-base to perform question and answer processing on the question and answer data. The first prompt word is input into the large language model for the second question-and-answer process to obtain the second answer data.

6. The control method for the intelligent toy as described in claim 4, characterized in that, The second question-and-answer processing, based on the target general knowledge base, on the user's question-and-answer data to obtain second answer data, includes: Obtain the historical dialogue dataset between the user and the smart toy, the historical dialogue dataset containing at least one historical question-and-answer pair; Based on the target knowledge sub-base, the historical dialogue dataset, and the user's question-and-answer data, a second prompt word for the large language model is constructed. The second prompt word is used to instruct the large language model to call the target knowledge sub-base to perform question-and-answer processing on the question-and-answer data. The second prompt word is input into the large language model for the second question-and-answer process to obtain the second answer data.

7. The control method for the intelligent toy as described in claim 6, characterized in that, After performing second question-and-answer processing on the user's question-and-answer data based on the target general knowledge base to obtain second answer data, and controlling the smart toy to output the second answer data, the method further includes: If there is no target general knowledge base corresponding to the user's age information, a temporary knowledge base is generated based on the online knowledge platform; Based on the temporary knowledge base, the user's question and answer data is processed in a third question and answer manner to obtain third answer data, and the smart toy is controlled to output the third answer data.

8. The control method for the intelligent toy as described in claim 7, characterized in that, The process of performing third question-and-answer processing on the user's question-and-answer data to obtain third answer data, and controlling the smart toy to output the third answer data, includes: The user's question and answer data is subjected to third-party question and answer processing to obtain the answer data to be processed; The answer data to be processed is reviewed to determine whether the review of the answer data to be processed is successful. If the answer data to be processed is successfully reviewed, the answer data to be processed is determined to be the third answer data, and the smart toy is controlled to output the third answer data.

9. A control device for an intelligent toy, characterized in that, The control device for the intelligent toy includes: The first determining module is used to determine whether a target-specific knowledge base corresponding to the current growth value of the smart toy exists when the smart toy is in a question-and-answer state. The first question-and-answer processing module is used to, if there is a target-specific knowledge base corresponding to the current growth value, perform first question-and-answer processing on the user's question-and-answer data based on the target-specific knowledge base to obtain first answer data, and control the smart toy to output the first answer data; The second question-and-answer processing module is used to determine a target general knowledge base corresponding to the user's age information if there is no target knowledge base corresponding to the current growth value. The different levels of the general knowledge base correspond to different ages. The control module is used to perform second question-and-answer processing on the user's question-and-answer data based on the target general knowledge base, obtain second answer data, and control the smart toy to output the second answer data.

10. An electronic device, characterized in that, include: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the computer program, implements the steps of the control method for the intelligent toy as claimed in any one of claims 1 to 8.

11. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the steps of the control method for the intelligent toy as described in any one of claims 1 to 8.

Images