Interaction method and apparatus, and electronic device and storage medium
Patent Information
- Application Number
- PCT/CN2025/090935
- Authority / Receiving Office
- WO · WO
- Patent Type
- Applications
- Current Assignee / Owner
- Priority Date
- 2025-03-25
- Filing Date
- 2025-04-24
- Publication Date
- 2026-10-01
Smart Images
Figure CN2025090935_01102026_PF_FP_ABST
Abstract
Description
Interaction methods, devices, electronic devices and storage media
[0001] This application claims priority to Chinese Patent Application No. 2025103580562, filed on March 25, 2025, entitled “Interactive Method, Apparatus, Electronic Device and Storage Medium”, the entire contents of which are incorporated herein by reference. Technical Field
[0002] This application relates to the field of artificial intelligence technology, specifically to an interaction method, device, electronic device, and storage medium. Background Technology
[0003] With the development of artificial intelligence technology and the rise of large language models, the automotive industry is increasingly using this technology to meet user needs.
[0004] When users want to determine the vehicle information supported by a particular vehicle diagnostic product, such as the brand, model, year, and functions, they currently rely primarily on manually searching through the product's interface (which includes information on supported vehicles). However, due to the vast amount of data on supported vehicle brands and models, users encounter cumbersome search processes and incomplete information display during webpage searches. This not only results in low accuracy but also inefficiency, failing to meet users' needs for easily and efficiently obtaining supported vehicle information.
[0005] Therefore, how to efficiently and accurately determine the vehicle information that supports diagnosis in order to improve the user experience is an urgent problem to be solved. Summary of the Invention
[0006] This application provides an interaction method, apparatus, electronic device, and storage medium that can efficiently and accurately determine vehicle information that supports diagnostics, thereby improving the user experience.
[0007] In a first aspect, this application provides an interaction method applied to an interactive device, the method comprising:
[0008] In response to user input, retrieve the first question, which is used to query vehicle information that supports diagnostics;
[0009] The first answer is displayed, which is obtained based on the first question and the first model. The first model is trained using vehicle data corresponding to the vehicle supported by the vehicle diagnostic product.
[0010] Secondly, this application provides an interaction method applied to a server, the method comprising:
[0011] Obtain the first question, which is used to query vehicle information that supports the diagnosis;
[0012] Based on the first question and the first model, a first answer corresponding to the first question is generated. The first model is trained using vehicle data corresponding to vehicles that can be diagnosed with the support of vehicle diagnostic products.
[0013] Thirdly, this application provides an interactive device, which includes: a first acquisition unit and a first processing unit;
[0014] The first acquisition unit is used to respond to the user's input operation and acquire the first question, which is used to query vehicle information that supports diagnosis.
[0015] The first processing unit is used to display the first answer corresponding to the first question, wherein the first answer is obtained based on the first question and the first model, and the first model is trained by vehicle data corresponding to the vehicle supported by the vehicle diagnostic product.
[0016] Fourthly, this application provides a server, including: a second acquisition unit and a second processing unit;
[0017] The second acquisition unit is used to acquire the first question, which is used to query vehicle information that supports diagnosis.
[0018] The second processing unit is used to generate a first answer corresponding to the first question based on the first question and the first model. The first model is trained using vehicle data corresponding to vehicles that can be diagnosed by the vehicle diagnostic product.
[0019] Fifthly, this application provides an electronic device, including: a processor and a memory, the processor being connected to the memory, the memory being used to store a computer program, and the processor being used to execute the computer program stored in the memory, so that the electronic device performs the methods as described in the first and second aspects.
[0020] In a sixth aspect, this application provides a computer-readable storage medium storing a computer program, which, when executed by a processor, performs the methods of the first and second aspects.
[0021] In a seventh aspect, this application provides a computer program product, which includes a non-transitory computer-readable storage medium storing a computer program, wherein the computer program, when executed by a processor, performs the methods as described in the first and second aspects.
[0022] Implementing this application will have the following beneficial effects:
[0023] In the embodiments of this application, a first model is trained using vehicle data corresponding to vehicles supported for diagnosis by a vehicle diagnostic product. Then, the interactive device responds to the user's input operation and obtains a first question, which is used to query vehicle information that supports diagnosis. Then, based on the first question and the first model, a first answer corresponding to the first question is obtained. Then, the interactive device displays the first answer. That is, by using artificial intelligence, the user is helped to query or determine vehicle information that supports diagnosis without having to manually search for it on the interface of the vehicle diagnostic product. This is not only efficient and accurate, but also improves the user experience. Attached Figure Description
[0024] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0025] Figure 1 is a schematic diagram of the structure of a first model provided in an embodiment of this application;
[0026] Figure 2 is a schematic diagram of an interactive system provided in an embodiment of this application;
[0027] Figure 3 is a schematic diagram of the structure of a knowledge graph based on vehicle data supporting diagnostics provided in an embodiment of this application;
[0028] Figure 4 is a flowchart illustrating an interaction method provided in an embodiment of this application;
[0029] Figure 5 is a schematic diagram of determining user intent based on preset function buttons according to an embodiment of this application;
[0030] Figure 6 is a schematic diagram of the interaction process of an interaction method provided in an embodiment of this application;
[0031] Figure 7 illustrates a training method for a first model provided in an embodiment of this application;
[0032] Figure 8 is a schematic diagram of another first model provided in an embodiment of this application;
[0033] Figure 9 is a structural schematic diagram of another first model provided in an embodiment of this application;
[0034] Figure 10 is a schematic diagram of a scenario provided by an embodiment of this application;
[0035] Figure 11 is a schematic diagram of another scenario provided by an embodiment of this application;
[0036] Figure 12 is a block diagram of the functional units of an interactive device provided in an embodiment of this application;
[0037] Figure 13 is a block diagram of the functional units of a server provided in an embodiment of this application;
[0038] Figure 14 is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation
[0039] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.
[0040] The terms "first," "second," "third," and "fourth," etc., used in the specification, claims, and drawings of this application are used to distinguish different objects, not to describe a specific order. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or apparatus that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to these processes, methods, products, or apparatuses.
[0041] In this document, the term "embodiment" means that a particular feature, result, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.
[0042] First, the relevant terms and technologies involved in the embodiments of this application will be explained:
[0043] First model: In the embodiments of this application, the first model can be a large model based on the Transformer architecture, such as the GPT (Generative Pre-trained Transformer) model. The first model can mainly consist of multiple decoders in the Transformer architecture.
[0044] For example, refer to Figure 1, which is a schematic diagram of the structure of a first model provided in an embodiment of this application.
[0045] As shown in Figure 1, the first model includes N decoders. Each decoder mainly includes an embedding layer, a masked multi-head attention layer, a normalization layer (Norm), a feed-forward neural network, a linear layer (Linear), and an activation layer (Softmax). Taking a decoder as an example, for the input of the first model, the input is embedded and positionally encoded to obtain the first feature vector. Then, the first feature vector is processed by a masked multi-head attention mechanism to obtain the second feature vector corresponding to the input. Next, the first and second feature vectors are joined by a residual connection and then normalized to obtain the third feature vector. The third feature vector is then input into a feedforward neural network for processing to obtain the fourth feature vector. The third and fourth feature vectors are then joined by a residual connection and then normalized to obtain the fifth feature vector. Finally, the fifth feature vector is input into a linear layer for linear processing and activation processing to obtain the predicted probability corresponding to the input, typically a probability distribution. Based on this predicted probability, the final output can be determined. Specifically:
[0046] If it is a discriminative classification problem, such as binary classification or multi-class classification, then the predicted probability obtained by Softmax processing can be understood as the predicted probability of the input in each category, that is, a probability distribution. If it belongs to the category, the probability is 1, otherwise it is 0. The corresponding label is the true probability of the input in each category. Then, the category with the highest probability in the probability distribution corresponding to the input can be determined as the output corresponding to the input.
[0047] If it is a generative problem, such as the input is a question and the output is the corresponding answer, since generative problems are autoregressive, unlike discriminative problems which can be obtained by directly processing a probability distribution using Linear or Softmax to get a complete output, generative problems do not output a complete output sequence all at once, but rather predict and generate elements in the output column step by step until a complete output is obtained.
[0048] First, refer to Figure 2, which is a schematic diagram of an interactive system provided in an embodiment of this application.
[0049] As shown in Figure 2, the interactive system includes an interactive device and a server. The server can be a standalone physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server providing basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, and big data and artificial intelligence platforms. This application does not impose specific limitations. The interactive device can be a user terminal, such as a smartphone, tablet, laptop, desktop computer, smart TV, desktop computer, smartwatch, smart vehicle, etc., but is not limited to these. The number of user terminals can be one or more, which is also not limited in this application. The interactive device can have a target application or target webpage installed, allowing users to query and interact with data (such as querying vehicle information supporting diagnostics) through the target application or target webpage, thus realizing human-computer interaction. Specifically:
[0050] Users can perform data queries through target applications or web pages on the interactive device. For example, they can input information on the first interface of the target application or the target web page. The interactive device then responds to the user's input, obtaining the first question entered by the user. This first question represents the vehicle information that the user wants to query that supports diagnostics. The interactive device can then send the first question to the server. At this time, the server has a first model deployed on it. The first model is trained using vehicle data corresponding to the vehicles supported by the vehicle diagnostic product (it should be noted that the training method of the first model is not described in detail here; please refer to the corresponding explanation in the embodiments below). The server then inputs the first question into the first model and outputs a first answer corresponding to the first question. The server then sends the first answer to the interactive device, and the interactive device receives and displays the first answer. At this point, the first question and the first answer constitute a dialogue (also known as a one-turn dialogue). This process can be repeated, allowing users to conduct multiple rounds of dialogue through the interactive device, with each round following a similar principle.
[0051] It should be noted that the interactive device and server in the interactive system shown in Figure 2 can also perform other steps corresponding to those in the embodiments below, which will not be described in detail here. Please refer to the embodiments below for details.
[0052] It should be noted that users can query relevant data based on the interaction method of this application, such as querying vehicle information that supports diagnostics, querying vehicle repair plans, vehicle sales information, and any other questions they wish to ask. This application does not limit this; the embodiments below mainly use the intention of querying vehicle information that supports diagnostics as an example for explanation. Before introducing the method embodiments, the vehicle data corresponding to the vehicles supported by the vehicle diagnostic product involved in the embodiments of this application will be explained as follows:
[0053] In the embodiments of this application, vehicle data corresponding to vehicles supported by the vehicle diagnostic product can be collected. This data mainly includes modules such as the vehicle brand, vehicle model, and vehicle year supported by the diagnostic product. In addition, it may also include modules such as vehicle system and vehicle function, which will not be listed in this application. Vehicle types may include various types of vehicles such as commercial vehicles, passenger vehicles, motorcycles, and new energy vehicles, which will not be listed here. Each type of vehicle may contain different brands of vehicles, and vehicles under each brand may include different models, years, systems, functions, etc.
[0054] Furthermore, a corresponding first database can be constructed based on the vehicle data supported by these vehicle diagnostic products. At this time, each piece of data in the first database can be stored in a first preset format. All the data in the first database can fully reflect the correspondence or association between the contents of each module in the above-mentioned vehicle data, such as all models under a brand.
[0055] The following explanation uses the vehicle data, which includes modules such as vehicle brand, vehicle model, vehicle year, vehicle system, and vehicle function, as an example. For instance, the first preset format is "Vehicle Type_Vehicle Brand_Vehicle Model_Vehicle Year_Vehicle System_Vehicle Function". In this format, the contents of each module can be interchanged, and the number of contents belonging to the same module can be one or more, such as multiple vehicle models or multiple vehicle years. "_" represents a separator (other symbols can also be used as separators, which are not limited in this application). If some vehicles lack certain contents in vehicle type, vehicle brand, vehicle model, vehicle year, vehicle system, or vehicle function, then that portion of the data will be empty (NULL), ensuring that any two data entries are different. For example:
[0056] First data: Brand 1_Model 1_Year 1_System 1_Function 1;
[0057] Second data: Brand 1_Model 2_Year 2_System 1_Function 1;
[0058] Third data: Brand 1_Model 3_Year 3_System 1_NULL.
[0059] In an optional embodiment, the first database can also be a knowledge graph generated based on vehicle data. In this case, template content such as vehicle type, vehicle brand, vehicle signal, vehicle year, vehicle system, and vehicle function in the vehicle data can be used as entities in the knowledge graph. The correspondence or association between the template content such as vehicle type, vehicle brand, vehicle signal, vehicle year, vehicle system, and vehicle function is represented by connecting lines, and together they generate a knowledge graph corresponding to the vehicle data supporting diagnosis.
[0060] For example, referring to Figure 3, Figure 3 is a schematic diagram of the structure of a knowledge graph based on vehicle data supporting diagnostics provided in an embodiment of this application. As shown in Figure 3, each content in Figure 3 is an entity in the knowledge graph, and the connecting lines indicate the existence of association or correspondence. Specifically: the models corresponding to brand 1 include model 1-1, model 1-2, model 1-3, model 1-4, model 1-5, and model 1-6; the year corresponding to brand 1 and model 1-1 includes year 2, the year corresponding to brand 1 and model 1-2 includes year 2, the year corresponding to brand 1 and model 1-3 includes year 1, the year corresponding to brand 1 and model 1-4 includes year 3, and the corresponding system includes system 2, the year corresponding to brand 1 and model 1-5 includes year 3, and the corresponding system includes system 2, and the year corresponding to brand 1 and model 1-6 includes year 4, and the corresponding system includes system 1. Similarly, the models corresponding to brand 2 include model 2-1, model 2-2, model 2-3, and model 2-4; the year corresponding to brand 2 and model 2-1 includes year 5; the year corresponding to brand 2 and model 2-2 includes year 1; the year corresponding to brand 2 and model 2-3 includes year 3 and the corresponding system includes system 2; the year corresponding to brand 2 and model 2-4 includes year 5, the corresponding system includes system 3, and the corresponding function includes function 1.
[0061] Furthermore, referring to Figure 4, Figure 4 is a flowchart illustrating an interaction method provided in an embodiment of this application. This method is applied to the interaction device in the above embodiments; the method includes, but is not limited to, steps S401-S402:
[0062] S401, The interactive device responds to the user's input operation and obtains a first question, which is used to query vehicle information that supports diagnosis.
[0063] In embodiments of this application, a user can input a first question through an interactive device (such as a target application or a target webpage, which will not be described in detail here); then the interactive device responds to the user's input operation and obtains the first question input by the user, wherein the first question may be used to characterize querying vehicle information that supports diagnosis; then the interactive device sends the first question to the server.
[0064] S402. The interactive device displays the first answer corresponding to the first question, wherein the first answer is obtained based on the first question and the first model, and the first model is trained using vehicle data corresponding to the vehicle supported by the vehicle diagnostic product.
[0065] Accordingly, the server receives the first question from the interactive device, and then generates a first answer corresponding to the first question based on the first question and the first model. The server then sends the first answer to the interactive device, which receives and displays it for the user to view. Furthermore, the first question can be in text or audio format; this application does not limit this. When the first question is in audio format, the server can perform text conversion on the first question, such as using Automatic Speech Recognition (ASR) technology, and then execute subsequent steps based on the converted text. This will not be elaborated further here.
[0066] The server generates a first answer corresponding to the first question based on the first question and the first model, specifically including steps S11-S12:
[0067] S11. Obtain the user's first intent, which is used to determine vehicle information that supports the diagnostics.
[0068] In the embodiments of this application, the methods for obtaining the first intent may include at least the following two methods:
[0069] (1) After receiving the first question from the interactive device, the server performs keyword recognition on the first question to obtain the first keyword. The number of first keywords can be one or more, which is not limited in this application. Then, the server matches the first keyword with the preset keywords corresponding to each preset intent among multiple preset intents to determine the first intent from the multiple preset intents. For example, it calculates the similarity between the first keyword and the preset keywords corresponding to each preset intent. This application does not limit the specific principle of similarity determination. Then, the server determines the preset intent corresponding to the preset keyword with the highest similarity as the first intent. That is to say, in this method, the server does not receive the user's first intent from the interactive device, but determines the first intent based on the user's input first question.
[0070] (2) The first interface or target webpage of the target application of the interactive device may include at least one preset function button, each preset function button corresponding to an intent. For example, the intent may be to determine / query the vehicle information supported by the current vehicle diagnostic product, to obtain / query the repair plan for the vehicle, to query the after-sales rules corresponding to the product sold, etc. If the user wants to determine / query the supported vehicle information, he / she can touch or select the first function button; then the interactive device responds to the user's touch operation on the first function button and obtains the user's first intent; after obtaining the first question input by the user, the interactive device will send the first question and the first intent to the server, and the server will obtain the first question and the first intent accordingly. That is to say, in this method, the user's intent is that the user selects the first function button corresponding to the first intent through the interactive device, and then the interactive device directly sends the first intent to the server. Of course, if the user does not select the preset function button corresponding to the intent, it is still necessary to determine the user's intent based on the first method (1), which will not be elaborated here.
[0071] For ease of understanding, the following explanation is provided in conjunction with the accompanying drawings. Referring to Figure 5, Figure 5 is a schematic diagram of determining user intent based on preset function buttons according to an embodiment of this application. As shown in Figure 5, the interface shown in Figure 5 is the first interface or target webpage of the target application displayed by the interactive device, including an input box (for user input, such as a question), a function button corresponding to function 1, a first function button corresponding to function 2, a function button corresponding to function 3, and a send function button. Function 1, function 2, and function 3 each correspond to an intent, such as function 2 indicating confirming / querying supported vehicle information. The number of function buttons can be multiple; this application only uses three as an example for illustration. For example, when a user wants to confirm / query supported vehicle information, the user can select the first function button, enter the first question in the input box, and then click "send." The interactive device obtains the intent corresponding to the first function button, namely the first intent and the first question, and sends the first intent and the first question to the server. It should also be noted that there is no specific order between the user selecting the first function button and entering the first question in the input box; this application does not impose any restrictions on this. Of course, users can also choose not to select the corresponding first function button, but simply enter the first question in the input box and then click "send". The interactive device will then obtain the first question and send it to the server, which will determine the first intent based on the first question.
[0072] S12. Based on the first question, the database corresponding to the first intent, and the first model, obtain the first answer.
[0073] For example, the server can first retrieve first data related to the first question from a first database corresponding to the first intent, based on the first question. The first database is generated based on vehicle data corresponding to vehicles supported by the vehicle diagnostic product. In the embodiments of this application, a database corresponding to each intent can be pre-constructed. For example, a first database corresponding to the first intent can be constructed, the principle of which will not be elaborated here. Another example is the intent to obtain / query a repair solution for a vehicle, which can be based on various repair data corresponding to the vehicle to construct a database corresponding to that intent. Then, the server retrieves first data related to the first question from the first database corresponding to the first intent, based on the first question. For example, by calculating the similarity between the first question and the data in the first database (this application does not limit the specific principle); then, data with a similarity greater than a first threshold is used as the first data.
[0074] Then, the server generates a first prompt word based on the first question and the first data. For example, the server can concatenate the first question and the first data as the first prompt word; or, the server can embed the first question and the first data into a preset prompt word template to obtain the first prompt word.
[0075] Then, the server obtains the first answer based on the first prompt word and the first model. For example, the server can input the first prompt word into the first model, that is, use the first prompt word as the input of the first model and output the first answer corresponding to the first question. In this case, the specific steps performed by using the first prompt word as the input of the first model can be referred to the steps performed by the first model on the input in the embodiment of Figure 1 above, which will not be repeated here.
[0076] In one optional embodiment, regarding obtaining the first answer based on the first question, a database corresponding to the first intent, and a first model, after obtaining the first intent, the server can generate multiple third answers based on the first database and multiple preset answer templates, with each third answer corresponding to one of the preset answer templates. Then, based on the first question, the server determines the first matching result corresponding to the first question from the multiple third answers. For example, it can calculate the similarity between the first question and each third answer, and then use the third answers with a similarity greater than a second threshold as the first matching result. Then, the server generates the first answer based on the first matching result, the first database, the first question, and the first model. Specifically:
[0077] At this point, the first database is a knowledge graph generated based on vehicle data. If the first matching result is empty, the server annotates the first question according to the entities in the knowledge graph (details omitted here) to obtain the fourth question. For example, it can identify the third entity of the first question. If the knowledge graph includes the third entity, the third entity in the first question is annotated. There are many forms of annotation, such as using separators to annotate the third entity; this application does not impose any specific limitations. Then, based on the fourth question, the knowledge graph, and the first model, the server generates the first answer. That is, the server uses the fourth question as a prompt word and inputs it into the first model along with the knowledge graph. Then, the server executes the following steps through the first model:
[0078] First, feature extraction is performed on the fourth problem to obtain the ninth feature, which can be understood as the steps corresponding to the embedding layer and positional encoding in Figure 1. Then, feature extraction is performed on the knowledge graph to obtain the second feature corresponding to the knowledge graph. For example, the entities and relationships between entities in the knowledge graph can be converted into the corresponding graph structure. At this time, the entities correspond to the nodes in the graph structure, and the relationships between entities correspond to the edges in the graph structure. Then, feature extraction is performed on the graph structure to obtain the second feature. The second feature can be understood as a feature matrix, which includes the feature vector corresponding to each entity in the knowledge graph (or the feature vector corresponding to each node in the graph structure).
[0079] It should be noted that this application can use a matching feature extractor to extract features from the fourth question and the knowledge graph respectively. For example, a first feature extractor can be used to extract features from the fourth question (as shown in Figure 1, corresponding to the embedding layer and position encoding), and a second feature extractor can be used to extract features from the knowledge graph. The second feature extractor can be an encoder used to extract features from an image.
[0080] Then, the third feature is obtained by performing graph convolution on the second feature. For example, graph convolution can be performed on the second feature using a Graph Convolutional Network (GCN). The specific principle of graph convolution is not elaborated here, but is explained in the corresponding embodiments below. Then, attention processing is performed on the ninth and third features to obtain the tenth feature. For example, cross-attention processing can be performed on the ninth and third features, or self-attention processing can be performed on the ninth and third features separately before cross-attention processing. This application does not limit this. Finally, a second answer is generated based on the tenth feature. For example, the fourth feature is processed according to the steps performed by the first model on the first feature vector in the embodiment of Figure 1 to obtain the first answer. Specifically:
[0081] The tenth feature is processed using a masked multi-head attention mechanism to obtain the eleventh feature. Then, the tenth and eleventh feature vectors are connected by a residual connection and normalized to obtain the twelfth feature corresponding to the input. The twelfth feature is then input into a feedforward neural network for processing to obtain the thirteenth feature. The twelfth and thirteenth features are then connected by a residual connection and normalized (Norm) to obtain the fourteenth feature. The fourteenth feature is then input into a linear layer for linear processing and activation processing to obtain the prediction probability corresponding to the first question. Based on this prediction probability and the first prediction probability, the first answer can be obtained.
[0082] In one optional embodiment, if the user does not know the specific vehicle information, the user can input a first vehicle image and a first question corresponding to the first vehicle image. The first question can be used to query whether the vehicle (or vehicle information) in the first vehicle image supports diagnosis. The interactive device responds to the user's input operation, obtains the first vehicle image and the first question input by the user, and sends the first question and the first vehicle image to the server.
[0083] At this point, the diagnostic product supports vehicle data including text data and image data. The text data includes one or more of the vehicle's brand, model, year, system, and function, while the image data includes images of the vehicle. The server then generates multiple image-text pairs based on the text and image data. Within each image-text pair, there is a relationship between the text and the image; for example, the vehicle information (such as brand) in the image corresponds to the text (such as model). This will not be elaborated further here. The server then executes the following steps using the first model:
[0084] For each image-text pair, feature extraction is performed on the image to obtain the first image feature corresponding to the image in each image-text pair, and feature extraction is performed on the text in each image-text pair to obtain the fifth feature corresponding to the text in each image-text pair. Then, attention processing is applied to the first image feature and the fifth feature corresponding to each image-text pair to obtain the sixth feature corresponding to each image-text pair. Of course, before performing attention processing, the first image feature and the fifth feature can be mapped to the same dimensional space.
[0085] Then, feature extraction is performed on the first vehicle image to obtain the third image feature. Then, based on the sixth, third, and tenth features corresponding to each image-text pair (the principle is not elaborated here), the first answer is generated. For example, the sixth, third, and fourth features corresponding to each image-text pair can be fused (e.g., spliced, weighted) to obtain the first fused feature. Alternatively, the server can first determine the first similarity between the third image feature and the sixth feature corresponding to each image-text pair; then, based on the first similarity, the server filters multiple image-text pairs to obtain the first image-text pair; then, the sixth and tenth features corresponding to the first image-text pair are fused to obtain the first fused feature.
[0086] Then the server uses the first model to predict the first fused feature and generate the first answer. For example, the first fused feature is processed according to the steps of the first model on the first feature vector in the embodiment of Figure 1 to obtain the first answer. The specific principle will not be elaborated here.
[0087] Conversely, if the first matching result is not empty, the server performs entity recognition on the first problem to obtain the third entity. Then, based on the third entity and the vehicle data supported by the aforementioned vehicle diagnostic product, the server determines the fourth entity that is associated with the third entity. For example, entities that are directly or indirectly connected in the embodiment of Figure 3 above are associated. Then, the server obtains the historical diagnostic data of the vehicle diagnostic product for the third entity and the fourth entity. For example, if the third entity and the fourth entity correspond to "Brand 1_Model 1", then the historical diagnostic data is the diagnostic data for vehicles of "Brand 1_Model 1" (such as the total number of diagnoses, the number of systems diagnosed, the functions and number of diagnoses, the quality feedback results of diagnoses, etc.).
[0088] The server then uses the first model to extract features from historical diagnostic data, obtaining the seventh feature, and from the first matching result, obtaining the fifteenth feature. Based on the seventh and fifteenth features, the server then uses the first model to generate a first answer. For example, the seventh and fifteenth features are fused (details omitted here) to obtain a second fused feature. The first model then uses this second fused feature to make a prediction and generate the first answer. For instance, the second fused feature is processed according to the steps performed on the first feature vector by the first model in the embodiment shown in Figure 1 to obtain the first answer; the specific principles are not detailed here. The resulting first answer is an adjusted and optimized version of the first matching result. For example, the first answer includes not only the answer to the first question but also diagnostic reference data. This diagnostic reference data is generated based on historical diagnostic data, which not only enriches the answer content but also provides diagnostic reference for users, increasing their willingness to choose the diagnostic product.
[0089] The following is a flowchart illustrating another interaction method of this application. This method is applied to the server in the above embodiments; the method includes, but is not limited to, steps S21-S22:
[0090] S21. The server retrieves the first question, which is used to query vehicle information that supports the diagnosis.
[0091] In the embodiments of this application, a user can input a first question through an interactive device, and then the interactive device responds to the user's input operation by obtaining the first question, and then the interactive device sends the first question to the server, and correspondingly, the server receives the first question from the interactive device.
[0092] S22. Based on the first question and the first model, the server generates the first answer corresponding to the first question. The first model is trained using vehicle data corresponding to vehicles that can be diagnosed by the vehicle diagnostic product.
[0093] It should be noted that the specific principles of steps S21-S22 can be referred to the corresponding explanations in steps S401-S402 of the above embodiments, and will not be repeated here.
[0094] Referring to Figure 6, Figure 6 is a schematic diagram of the interaction flow of an interaction method provided in an embodiment of this application. This method is applied to the interaction system in the above embodiments; the method includes, but is not limited to, steps S601-S605:
[0095] S601, The interactive device responds to the user's input operation and obtains the first question.
[0096] The first question is used to query vehicle information that supports diagnostics.
[0097] S602, The interactive device sends the first question to the server.
[0098] S603. The server generates the first answer corresponding to the first question based on the first question and the first model.
[0099] The first model is trained using vehicle data corresponding to vehicles that are supported by vehicle diagnostic products.
[0100] S604, The server sends the first response to the interactive device.
[0101] S605, The interactive device displays the first answer.
[0102] It should be noted that the specific principles of steps S601-S605 can be referred to the corresponding explanations in the above embodiments, and will not be repeated here.
[0103] It should be noted that the first model in this application is trained using vehicle data corresponding to vehicles that can be diagnosed with the support of the vehicle diagnostic product. The training principle of the first model is explained below with reference to specific embodiments:
[0104] First, referring to the structure of the first model in the embodiment of Figure 1, a training method for the first model is introduced. Referring to Figure 7, Figure 7 shows a training method for the first model provided in this application embodiment. This method is applied to a server; the method includes, but is not limited to, steps S701-S705:
[0105] S701, Obtain the second question.
[0106] In the embodiments of this application, the second problem is used as a training sample. The number of training samples is not limited. The embodiments of this application mainly use one as an example for explanation.
[0107] S702. Generate a first database based on the vehicle data corresponding to the vehicles that can be diagnosed by the vehicle diagnostic product.
[0108] The production principle of the first database will not be elaborated here; please refer to the corresponding description in the above embodiments for details.
[0109] S703. Based on the second question and the first database, generate a second answer corresponding to the second question.
[0110] For example, firstly, based on the second question, a search is performed in the first database to obtain the second data corresponding to the second question. The principle is similar to that of obtaining the first data, and will not be repeated here. Then, based on the second question and the second data, a second prompt word is generated. The principle is similar to that of generating the first prompt word, and will not be repeated here. Then, based on the second prompt word, a second answer is generated. For example, the second prompt word is input into the first model, that is, the second prompt word is used as the input of the first model, and the second answer corresponding to the second question is output. At this time, the specific steps performed by using the second prompt word as the input of the first model can refer to the steps performed by the first model on the input in the embodiment of Figure 1 above, and will not be repeated here.
[0111] In an optional embodiment, referring to the embodiment in Figure 1 and Figure 8, Figure 8 is a schematic diagram of another first model provided by the embodiment of this application. As shown in Figure 8, the first model includes N layers. Taking one layer as an example, it mainly includes a feature extraction layer, graph convolutional networks, cross-attention layers, masked multi-head attention layers, a feed-forward neural network normalization layer (Norm), a linear layer, and an activation layer (Softmax). Among them, the feature extraction layer includes a first feature extractor (i.e., Feature Extraction (1) in Figure 8) and a second feature extractor (i.e., Feature Extraction (2) in Figure 8). The first feature extractor can be a text encoder, and the second feature extractor can be an image encoder. This application does not limit the scope of the feature extractor.
[0112] Therefore, in step S703, regarding generating a second answer corresponding to the second question based on the second question and the first database, the server can also generate multiple third answers based on the first database and multiple preset answer templates. These multiple third answers correspond one-to-one with the multiple preset answer templates; details will not be elaborated here. Then, based on the second question, the server determines the matching result corresponding to the second question from the multiple third answers. The principle is similar to that of the first matching result described above; details will not be elaborated here. Then, based on the matching result, the first database, and the second question, the server generates a second answer. If the matching result is empty, then specifically:
[0113] First, the server labels the second question with entities according to the entities in the knowledge graph to obtain the third question; then the third question is input into the first feature extractor Feature Extraction(1) for feature extraction to obtain the first feature, and the knowledge graph (or the corresponding graph structure) is input into the second feature extractor Feature Extraction(2) for feature extraction to obtain the second feature corresponding to the knowledge graph, which will not be elaborated further.
[0114] The second feature is then input into a Graph Convolutional Network for graph convolution processing to obtain the third feature. The first and third features are then input into a Cross-Attention Layer for cross-attention processing to obtain the fourth feature. A second answer is then generated based on the fourth feature, similar to the principle of generating a second answer based on the tenth feature. Alternatively, the first model in the embodiment shown in Figure 1 can perform the corresponding steps on the first feature vector to obtain the output, which will not be elaborated here. When generating the second answer based on the fourth feature, a residual connection (Add) and normalization (Norm) can be performed on the fourth, first, and third features to obtain the sixteenth feature. The second answer is then generated based on the sixteenth feature, i.e., the processing follows the corresponding steps performed by the first model on the first feature vector in the embodiment shown in Figure 1.
[0115] In one optional embodiment, based on the embodiment of FIG9, referring to FIG9, FIG9 is a schematic diagram of the structure of another first model provided by the embodiment of this application.
[0116] The feature extraction layer in the first model shown in Figure 9 also includes a third feature extractor, Feature Extraction (3), which includes a text feature extractor and an image feature extractor. The explanations of the remaining models refer to the explanation in the embodiment of Figure 8, and will not be repeated here. At this time, the aforementioned vehicle data may include text data and image data. The text data includes one or more of the vehicle's brand, model, year, system, and function that support diagnosis, and the image data includes vehicle images that support diagnosis. Accordingly, regarding the generation of the second answer based on the fourth feature,
[0117] The server generates multiple image-text pairs based on text and image data. There is a relationship between the text and the image in each image-text pair, which will not be elaborated here. Then, the image feature extractor in the third feature extractor of the first model extracts features from the image in each image-text pair to obtain the first image feature corresponding to the image in each image-text pair. The text feature extractor in the third feature extractor of the first model extracts features from the text in each image-text pair to obtain the fifth feature corresponding to the text in each image-text pair.
[0118] Then, the first and fifth image features corresponding to each image-text pair are input into the Cross-Attention Layer for cross-attention processing to obtain the sixth feature corresponding to each image-text pair. The server then acquires the first vehicle image corresponding to the first question and inputs it into the image feature extractor in the third feature extractor for feature extraction to obtain the second image feature.
[0119] Then, based on the sixth feature, second image feature, and fourth feature corresponding to each image-text pair, a second answer is generated, as shown in Figure 9. The sixth feature, second image feature, and fourth feature corresponding to each image-text pair are fused to obtain a third fused feature. Alternatively, multiple image-text pairs can be filtered based on the second image feature and the sixth feature corresponding to each image-text pair to obtain a first image-text pair. The principle will not be elaborated here. Then, the sixth feature and fourth feature corresponding to the first image-text pair are fused to obtain a third fused feature. This application does not limit this.
[0120] Then, the third fusion feature, the first image feature, the fifth feature, the first feature and the third feature corresponding to each image text pair are subjected to residual connection Add and normalization Norm processing to obtain the seventeenth feature; then, the second answer is generated based on the seventeenth feature, the principle of which is similar to the principle of generating the second answer based on the sixteenth feature, or in other words, the corresponding steps of the first model performing on the first feature vector in the embodiment of Figure 1 are processed, which will not be repeated here.
[0121] Conversely, if the matching result is not empty, the server, in generating the second answer based on the matching result, the first database, and the second question, first performs entity recognition on the second question to obtain the first entity; then, based on the first entity and vehicle data, it determines the second entity that is associated with the first entity (details omitted); then, it obtains historical diagnostic data of the vehicle diagnostic product for the first and second entities (details omitted); then, it extracts features from the historical diagnostic data through the first feature extractor in the first model to obtain the seventh feature, and extracts features from the matching result through the first feature extractor in the first model to obtain the eighth feature; then, based on the seventh and eighth features, it obtains the second answer, for example, by inputting the seventh and eighth features into the cross-attention layer for attention processing to obtain the eighteenth feature; then, it performs residual connection Add and normalization Norm processing on the eighteenth, seventh, and eighth features to obtain the nineteenth feature; then, it generates the second answer based on the nineteenth feature, similar to the principle of generating the second answer based on the sixteenth feature, or in other words, it processes the first feature vector according to the corresponding steps performed by the first model in the embodiment of Figure 1 (details omitted).
[0122] S704. Based on the second answer, determine the training loss.
[0123] After generating the second answer, loss calculation can be performed based on the second answer and the real labels corresponding to the training samples, i.e., the second question. For example, cross-entropy loss, mean squared error loss, etc. This application does not limit the type of loss, and the training loss is obtained.
[0124] S705. Train the first model based on the training loss.
[0125] Then, the model can be trained based on the training loss. For example, the first model can be fine-tuned with all parameters, or with some parameters, such as the weights in the masked multi-head attention or the weights in the cross attention. For example, LoRA fine-tuning can be used. This application does not limit the training until the first model converges and the trained first model is obtained.
[0126] Furthermore, after training the first model according to the above embodiments until the first model converges, applications can be applied based on the trained first model. The main application scenarios involved in this application are described below with reference to the accompanying drawings. This scenario includes a server and an interactive device, with the first model deployed on the server, as detailed below:
[0127] Referring to the embodiment in Figure 5 and Figure 10, Figure 10 is a schematic diagram of a scenario provided by an embodiment of this application. As shown in Figure 10, the user enters the question "Does the 2020 BMW X5 support diagnostics?" in the input box displayed on the first interface of the target application or the target webpage on the interactive device. At this time, the question is used to query the vehicle information that supports diagnostics, namely, the brand is BMW, the model is X5, and the year is 2020. The user selects the first function button corresponding to function 2 (at this time, the function button is in the selected state, as shown by the shaded filling in Figure 10), and the corresponding intent of the first function button is "to determine the vehicle information that supports diagnostics". Then the user can click the "send" function button, and the interactive device responds to this operation by sending the question and intent to the server.
[0128] Accordingly, the interactive device can display the user's input question (not shown in Figure 10), and the server receives the question and intent. Then, the server generates an answer corresponding to the question based on the first model, the question, and the database corresponding to the intent. The specific principle will not be elaborated here. An example answer shown in Figure 10 is: "For your BMW X5 model, our device can diagnose models from 1998 to 2024. Therefore, the 2020 BMW X5 is diagnosable. In addition to the X5, we also support diagnosing other BMW models, such as the 1 Series, 2 Series, 3 Series, etc., totaling approximately 29 models. If you have questions about other models or years, please feel free to ask." The server then sends this answer to the interactive device, which receives and displays the answer, presenting a dialogue as shown in Figure 10.
[0129] Of course, this is optional. Referring to Figure 11, which is a schematic diagram of another scenario provided by an embodiment of this application, as shown in Figure 11, the user can also choose not to select a function button and simply enter the question "Does the 2020 BMW X5 support diagnostics?" in the input box. In this case, the question is used to query vehicle information that supports diagnostics, namely, the brand is BMW, the model is X5, and the year is 2020. Then the user can click the "Send" function button, and the interactive device responds to this operation by sending the question to the server.
[0130] Accordingly, the interactive device can display the user-inputted question (not shown in Figure 11), and the server receives the question; then the server determines the corresponding intent based on the question; then the server generates an answer corresponding to the question through the first model, based on the question and the database corresponding to the intent. The specific principle will not be elaborated here. The example answer shown in Figure 11 is: "According to professional diagnostic knowledge, the BMW X5 supports model years from 1998 to 2024. Therefore, the 2020 BMW X5 supports diagnostics. In addition to the BMW X5, there are many other models and supported model years in the BMW series, such as the 1 Series (2003-2024), 3 Series (1996-2024), 5 Series (1996-2024), etc. There are approximately 29 models." The server then sends this answer to the interactive device, and the interactive device receives and displays the answer, presenting a round of dialogue as shown in Figure 11.
[0131] It should be noted that the content shown in the embodiments of Figures 10 and 11, such as questions, answers, and page layouts, are merely examples. Any extensions or variations based on these examples, as well as the application of the interaction methods of this application in other scenarios, are all within the scope of protection of this application.
[0132] Referring to Figure 12, Figure 12 is a functional unit block diagram of an interactive device provided in an embodiment of this application. The interactive device 1200 includes: a first acquisition unit 1201 and a first processing unit 1202;
[0133] The first acquisition unit 1201 is used to acquire a first question in response to the user's input operation. The first question is used to query vehicle information that supports diagnosis.
[0134] The first processing unit 1202 is used to display the first answer corresponding to the first question, wherein the first answer is obtained based on the first question and the first model, and the first model is trained by vehicle data corresponding to the vehicle supported by the vehicle diagnostic product.
[0135] In specific implementations, the first acquisition unit 1201 and the first processing unit 1202 described in the embodiments of the present invention may also execute other methods executed by the interactive device in other interactive method embodiments provided in the embodiments of the present invention, which will not be repeated here.
[0136] Referring to Figure 13, Figure 13 is a functional unit block diagram of a server provided in an embodiment of this application. The server 1300 includes: a second acquisition unit 1301 and a second processing unit 1302;
[0137] The second acquisition unit 1301 is used to acquire the first question, which is used to query vehicle information that supports diagnosis.
[0138] The second processing unit 1302 is used to generate a first answer corresponding to the first question based on the first question and the first model. The first model is trained using vehicle data corresponding to vehicles that can be diagnosed by the vehicle diagnostic product.
[0139] In specific implementations, the second acquisition unit 1301 and the second processing unit 1302 described in the embodiments of the present invention may also execute other methods executed by the server in the embodiments of the interaction method provided by the present invention, which will not be elaborated here.
[0140] Referring to Figure 14, which is a schematic diagram of the structure of an electronic device provided in an embodiment of this application, the electronic device 1400 includes a transceiver 1401, a processor 1402, and a memory 1403. These components are connected via a bus 1404. The memory 1403 stores computer programs and data, and can transmit data stored in the memory 1403 to the processor 1402.
[0141] Electronic device 1400 may be interactive device 1200 or server 1300;
[0142] When the electronic device 1400 is an interactive device 1200, the processor 1402 reads the computer program in the memory 1403 and performs the following operations:
[0143] The control transceiver 1401 responds to the user's input operation and obtains a first question, which is used to query vehicle information that supports diagnostics;
[0144] The first answer is displayed, which is obtained based on the first question and the first model. The first model is trained using vehicle data corresponding to the vehicle supported by the vehicle diagnostic product.
[0145] In specific implementations, the transceiver 1401 and processor 1402 described in the embodiments of the present invention can also execute other implementations described in the embodiments of the interaction method provided in the embodiments of the present invention, which will not be repeated here.
[0146] When electronic device 1400 is server 1300, processor 1402 reads computer program from memory 1403 and performs the following operations:
[0147] The control transceiver 1401 acquires the first question, which is used to query vehicle information that supports diagnostics;
[0148] Based on the first question and the first model, a first answer corresponding to the first question is generated. The first model is trained using vehicle data corresponding to vehicles that can be diagnosed with the support of vehicle diagnostic products.
[0149] In specific implementations, the transceiver 1401 and processor 1402 described in the embodiments of the present invention can also execute other implementations described in the embodiments of the interaction method provided in the embodiments of the present invention, which will not be repeated here.
[0150] Specifically, the transceiver 1401 may be the first acquisition unit 1201 of the interactive device 1200 in the embodiment of FIG12 or the second acquisition unit 1301 of the server 1300 in the embodiment of FIG13, and the processor 1402 may be the first processing unit 1202 of the interactive device 1200 in the embodiment of FIG12 or the second processing unit 1302 of the server 1300 in the embodiment of FIG13.
[0151] It should be understood that embodiments of this application also provide a computer-readable storage medium storing a computer program that is executed by a processor to implement some or all of the steps of any of the interactive methods described in the above method embodiments.
[0152] This application also provides a computer program product, which includes a non-transitory computer-readable storage medium storing a computer program operable to cause a computer to perform some or all of the steps of any of the interactive methods described in the above method embodiments.
[0153] It should be noted that, for the sake of simplicity, the foregoing method embodiments are all described as a series of actions. However, those skilled in the art should understand that this application is not limited to the described order of actions, as some steps may be performed in other orders or simultaneously according to this application. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are all optional embodiments, and the actions and modules involved are not necessarily essential to this application.
[0154] In the above embodiments, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions in other embodiments.
[0155] In the several embodiments provided in this application, it should be understood that the disclosed apparatus can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical or other forms.
[0156] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.
[0157] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software program module.
[0158] If the integrated unit is implemented as a software program module and sold or used as an independent product, it can be stored in a computer-readable storage device (CMD). Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a memory and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this application. The aforementioned memory includes various media capable of storing program code, such as a USB flash drive, read-only memory (ROM), random access memory (RAM), portable hard drive, magnetic disk, or optical disk.
[0159] Those skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, which may include: flash drive, read-only memory (ROM), random access memory (RAM), disk or optical disk, etc.
[0160] The embodiments of this application have been described in detail above. Specific examples have been used to illustrate the principles and implementation methods of this application. The description of the above embodiments is only for the purpose of helping to understand the method and core ideas of this application. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of this application. Therefore, the content of this specification should not be construed as a limitation of this application.
Claims
1. An interaction method, characterized in that, The method includes: In response to user input, a first question is retrieved, which is used to query vehicle information that supports diagnostics; The first answer corresponding to the first question is displayed, wherein the first answer is obtained based on the first question and the first model, and the first model is trained by vehicle data corresponding to the vehicle supported by the vehicle diagnostic product.
2. The method according to claim 1, characterized in that, Before displaying the first answer corresponding to the first question, the method further includes performing the following steps via a server: Obtain the user's first intent, which is used to determine vehicle information that supports the diagnostics; Based on the first question, the first database corresponding to the first intent, and the first model, the first answer is obtained. The first database is generated based on vehicle data corresponding to vehicles that are supported for diagnosis by the vehicle diagnostic product.
3. The method according to claim 2, characterized in that, The acquisition of the user's first intent includes: In response to the user's touch operation on the first function button, the user's first intention is obtained; or, Keyword identification was performed on the first question to obtain the first keyword; Based on the first keyword, the first intent is determined.
4. The method according to any one of claims 1-3, characterized in that, The training method for the first model includes the following steps: Get the second question; Based on the vehicle data, a first database is generated; Based on the second question and the first database, a second answer corresponding to the second question is generated; Based on the second answer, determine the training loss; The first model is trained based on the training loss.
5. The method according to claim 4, characterized in that, The step of generating a second answer corresponding to the second question based on the second question and the first database includes: Based on the first database and multiple preset response templates, multiple third responses are generated, and the multiple third responses correspond one-to-one with the multiple preset response templates; Based on the second question, determine the matching result corresponding to the second question from the plurality of third answers; Based on the matching results, the first database, and the second question, the second answer is generated.
6. The method according to claim 5, characterized in that, The first database is a knowledge graph generated based on the vehicle data; If the matching result is empty, then generating the second answer based on the matching result, the first database, and the second question includes: Based on the entities in the knowledge graph, the second question is labeled to obtain the third question; Feature extraction is performed on the third problem to obtain a first feature, and feature extraction is performed on the knowledge graph to obtain a second feature corresponding to the knowledge graph; The second feature is subjected to graph convolution to obtain the third feature; A fourth feature is obtained by performing attention processing based on the first feature and the third feature. The second answer is generated based on the fourth feature.
7. The method according to claim 5, characterized in that, If the matching result is not empty, then generating the second answer based on the matching result, the first database, and the second question includes: Entity recognition is performed on the second problem to obtain the first entity; Based on the first entity and the vehicle data, a second entity that is associated with the first entity is determined; Acquire the historical diagnostic data of the vehicle diagnostic product for the first entity and the second entity; Feature extraction is performed on the historical diagnostic data to obtain the seventh feature, and feature extraction is performed on the matching results to obtain the eighth feature; Based on the seventh feature and the eighth feature, the second answer is obtained.
8. An interactive device, characterized in that, The interactive device includes: a first acquisition unit and a first processing unit; The first acquisition unit is used to acquire a first question in response to the user's input operation, the first question being used to query vehicle information that supports diagnosis; The first processing unit is configured to display a first answer corresponding to the first question, wherein the first answer is obtained based on the first question and a first model, and the first model is trained using vehicle data corresponding to vehicles supported by a vehicle diagnostic product.
9. An electronic device, characterized in that, include: A processor and a memory, the processor being connected to the memory, the memory being used to store a computer program, and the processor being used to execute the computer program stored in the memory to cause the electronic device to perform the method as described in any one of claims 1-7.
10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that is executed by a processor to implement the method as described in any one of claims 1-7.