Product document generation method, electronic equipment and storage medium
By evaluating and optimizing the vehicle functional product documents, the quality of product documents has been greatly improved, the problem of low product documents in the existing technology has been solved, and more efficient document understanding and optimization has been achieved.
Patent Information
- Application Number
- CN202510253894.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-03-05
- Publication Date
- 2025-06-27
AI Technical Summary
The product documentation used in the prior art to describe vehicle functions is of low quality, resulting in difficulty in understanding users and maintenance personnel.
By obtaining user comments and competitor product documents from the network server, quality evaluation and comparison analysis of functional description text are carried out, weighted averages are carried out in combination with expert scores, and comprehensive scores are generated. Enter the function description text with a lower comprehensive score to the pre-trained text description optimization model for optimization, and generate a new product document based on the optimized text.
It improves the quality of product documents, enables users and maintenance personnel to fully understand the document content, optimizes it with high efficiency and saves labor costs.
Smart Images
Figure CN120218033A_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of vehicle technologies, and particularly to a method for generating product documents, an electronic device, and a storage medium. Background Art
[0002] Cockpit product documents refer to technical documents that describe information such as the functions, technical specifications, usage methods, maintenance, etc. of the cockpit products of a vehicle. The purpose of these technical documents is to provide comprehensive information for users and maintenance personnel to ensure the correct installation, operation, and maintenance of the products.
[0003] Currently, there may be some defects in the cockpit product documents when they are released and provided to users and maintenance personnel. For example, the functional descriptions are not detailed enough, making it difficult for users and maintenance personnel to understand; there are defects in the description of the operation steps of the functions, resulting in users and maintenance personnel being unable to implement the corresponding vehicle functions according to the functional description text in the product documents. Therefore, the quality of the currently released product documents for describing vehicle functions is not high. Summary of the Invention
[0004] This application provides a method for generating product documents, an electronic device, and a storage medium, which are used to solve the problem of low quality of product documents for describing vehicle functions in the prior art.
[0005] In a first aspect, this application provides a method for generating product documents, including:
[0006] Obtain the usage comments of a user on multiple vehicle functions integrated in a target application of a target vehicle model from a first network server; for each vehicle function, evaluate the quality of the functional description text of the vehicle function according to the corresponding usage comment to obtain a first score of the functional description text of the vehicle function;
[0007] Obtain the product documents of the target application of multiple competing vehicle models of the target vehicle model from a second network server, where the product documents of the target application of each competing vehicle model include the functional description texts of multiple vehicle functions;
[0008] For each vehicle function, compare and analyze the functional description text of the vehicle function in the target application of the preset target vehicle model with the functional description texts of the vehicle functions of each competing vehicle model to obtain a second score of the functional description text of the vehicle function;
[0009] For each vehicle function, receive a third score of the functional description text of the vehicle function in the target application of the preset target vehicle model from an expert terminal;
[0010] For each vehicle function, perform a weighted average on the first score, the second score, and the third score of the functional description text of the vehicle function to obtain a comprehensive score of the functional description text of the vehicle function;
[0011] Input the function description texts corresponding to the top N vehicle functions with relatively low comprehensive scores into a pre-trained large model for optimizing text descriptions to obtain N optimized function description texts. Among them, the large model for optimizing text descriptions is obtained by training multiple first training samples input into the large model network to be trained. Each first training sample includes the function description text before optimization and its corresponding historical comprehensive score in history, as well as the function description text after optimization and its corresponding historical comprehensive score in history. Here, N is an integer greater than or equal to 1;
[0012] Generate a new product document based on the N optimized function description texts and a preset product document template.
[0013] In some embodiments, perform quality assessment on the function description text of the vehicle function according to the corresponding usage comments to obtain the first score of the function description text of the vehicle function, including:
[0014] For each vehicle function, extract keywords associated with the usage problems from the usage comments corresponding to the vehicle function;
[0015] Perform quality assessment on the function description text of the vehicle function according to the keywords associated with the usage problems to obtain the first score of the function description text of the vehicle function.
[0016] In some embodiments, perform quality assessment on the function description text of the vehicle function according to the keywords associated with the usage problems to obtain the first score of the function description text of the vehicle function, including:
[0017] According to the keywords associated with the usage problems, determine the dissatisfaction sentiment index of the user with respect to the usage problems and the influence index of the usage problems on the normal driving of the vehicle;
[0018] Obtain the first score of the function description text of the vehicle function according to the dissatisfaction sentiment index and the influence index on the normal driving of the vehicle.
[0019] In some embodiments, obtain the first score of the function description text of the vehicle function according to the dissatisfaction sentiment index and the influence index on the normal driving of the vehicle, including:
[0020] Look up the sub-score corresponding to the dissatisfaction sentiment index in a preset first mapping relation table;
[0021] Look up the sub-score corresponding to the influence index on the normal driving of the vehicle in a preset second mapping relation table;
[0022] Perform weighted averaging on the sub-score corresponding to the dissatisfaction sentiment index and the sub-score corresponding to the influence index on the normal driving of the vehicle to obtain the first score of the function description text of the vehicle function.
[0023] In some embodiments, before obtaining the first score of the functional description text of the vehicle function according to the dissatisfaction emotion index and the influence index on the normal driving of the vehicle, the method provided by the present application further includes:
[0024] Obtain the operation behavior data of the vehicle function of the target vehicle model from the vehicle supplier server, where the operation behavior data at least includes the operation duration and the operation process steps;
[0025] Determine the operation experience index of the vehicle function of the target vehicle model according to the operation behavior data;
[0026] Obtain the first score of the functional description text of the vehicle function according to the dissatisfaction emotion index, the influence index on the normal driving of the vehicle, and the operation experience index.
[0027] In some embodiments, for each vehicle function, compare and analyze the functional description text of the vehicle function in the target application of the preset target vehicle model with the functional description texts of the vehicle functions of each competing vehicle to obtain the second score of the functional description text of the vehicle function, including:
[0028] For each vehicle function, input the functional description text of the vehicle function in the target application of the preset target vehicle model and the functional description texts of the vehicle functions of each competing vehicle into a pre-trained text comparison and analysis model to output the second score of the functional description text of the vehicle function, where the text comparison and analysis model is obtained by training a neural network to be trained with a plurality of second training samples, and each second training sample includes the historical functional description text of the historical vehicle function of each historical competing vehicle of the historical vehicle model, the historical functional description text of the historical vehicle function in the historical target application of the historical vehicle model, and the corresponding historical second score.
[0029] In some embodiments, the target vehicle model further includes: an associated application of the target application, and the associated application is associated with the functional description text corresponding to at least one vehicle function among the top N vehicle functions with relatively low comprehensive scores. After generating a new product document according to the optimized N functional description texts and the preset product document template, the method provided by the present application further includes:
[0030] Synchronously update the functional description text corresponding to at least one vehicle function associated with the associated application according to the optimized functional description text corresponding to at least one vehicle function.
[0031] In a second aspect, the present application further provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the electronic device is enabled to execute the method provided in the first aspect of the present application.
[0032] In a third aspect, the present application further provides a storage medium storing a computer program, which when executed by a processor causes the computer to execute the method provided in the first aspect of the present application.
[0033] In a fourth aspect, the present application further provides a computer program product including a computer program, which when run causes an electronic device to execute the method performed in the first aspect of the present application.
[0034] The present application provides a product document generation method, an electronic device, and a storage medium, which can perform a weighted average on the first score, the second score, and the third score of the function description text of each vehicle function to obtain a comprehensive score of the function description text of the vehicle function.
[0035] The function description texts corresponding to the top N vehicle functions with lower comprehensive scores are input into a pre-trained large model for optimizing text descriptions to obtain N optimized function description texts. Since the large model for optimizing text descriptions is obtained by training a plurality of first training samples input into a large model network to be trained, and each first training sample includes the function description text before optimization in history and its corresponding historical comprehensive score, as well as the function description text after optimization in history and its corresponding historical comprehensive score. In this way, the function description texts corresponding to the top N vehicle functions with lower comprehensive scores are optimized. Furthermore, according to the N optimized function description texts and a preset product document template, a new product document is generated. In this way, the quality of the new product document is improved, enabling users and maintenance personnel to fully understand the new product document, and the efficiency of optimizing the product document is high, saving labor costs. BRIEF DESCRIPTION OF THE DRAWINGS
[0036] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present application. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.
[0037] Figure 1 Schematic diagram of the interaction between the background server provided in the embodiment of the present application and the first network server, the second network server, and the expert terminal respectively;
[0038] Figure 2 Flowchart of the product document generation method provided in the embodiment of the present application;
[0039] Figure 3 Functional module block diagram of the product document generation device provided in the embodiment of the present application. Detailed Implementation Modes
[0040] Hereinafter, embodiments of the present disclosure will be described with reference to the accompanying drawings. However, it should be understood that these descriptions are merely exemplary and are not intended to limit the scope of the present disclosure. In addition, in the following description, descriptions of well-known structures and technologies are omitted to avoid unnecessarily confusing the concepts of the present disclosure.
[0041] Schematic diagrams of various structures according to embodiments of the present disclosure are shown in the accompanying drawings. These figures are not drawn to scale, where for the purpose of clear expression, some details are enlarged and some details may be omitted. The shapes of various regions and layers shown in the figures, as well as their relative sizes and positional relationships, are merely exemplary and may actually deviate due to manufacturing tolerances or technical limitations. Those skilled in the art can additionally design regions / layers with different shapes, sizes, and relative positions according to actual needs.
[0042] In the context of the present disclosure, when a layer / element is referred to as being "on" another layer / element, the layer / element may be directly on the other layer / element, or there may be an intermediate layer / element between them. Additionally, if a layer / element is "on" another layer / element in one orientation, then when the orientation is reversed, the layer / element may be "under" the other layer / element.
[0043] Hereinafter, the technical solutions of the present application and how the technical solutions of the present application solve the above technical problems will be described in detail with specific embodiments. These several specific embodiments below can be combined with each other, and the same or similar concepts or processes may not be repeated in some embodiments. The embodiments of the present application will be described below with reference to the accompanying drawings.
[0044] An embodiment of the present application provides a method for generating product documentation, which is applied to the back-end server 101. As Figure 1 shown, the back-end server 101 is communicatively connected to the first network server 102, the second network server 103, and the expert terminal 104 respectively. Among them, the first network server 102 may be, but is not limited to, the server of an automotive network dedicated to vehicle buying, selling, and introduction, and the second network server 103 may be the server of vehicle suppliers of various vehicle brands.
[0045] As Figure 2 shown, the method provided by the embodiment of the present application includes:
[0046] S201: Obtain usage comments of multiple vehicle functions integrated with the target application of the target vehicle model from the first network server 102.
[0047] Exemplarily, the target application can be an application associated with the cockpit function of a vehicle, or an application associated with the intelligent driving function, etc., which is not limited herein. For example, the target application can be a navigation application. The multiple vehicle functions integrated in the navigation application can include an address search function, a voice playback function, a route planning function, etc., which is not limited herein. The usage comment of the vehicle function can be "The address search of the navigation function is not very convenient, and the result is inaccurate after entering the place name". In addition, the target application can also be a music application, a phone application, an instrument application, etc., which is not limited herein.
[0048] S202: For each vehicle function, perform a quality assessment on the function description text of the vehicle function according to the corresponding usage comment to obtain the first score of the function description text of the vehicle function.
[0049] Specifically, for each vehicle function, extract the keywords associated with the usage problems from the usage comments corresponding to the vehicle function; according to the keywords associated with the usage problems, perform a quality assessment on the function description text of the vehicle function to obtain the first score of the function description text of the vehicle function.
[0050] Exemplarily, according to the keywords associated with the usage problems, the dissatisfaction emotion index of the user with respect to the usage problems and the influence index of the usage problems on the normal driving of the vehicle can be determined.
[0051] It should be noted that the higher the dissatisfaction emotion index, the higher the dissatisfaction degree of the user; the higher the influence index on the normal driving of the vehicle, the greater the influence on the normal driving of the vehicle.
[0052] For example, when the usage comment is "The address search of the navigation function is very inconvenient, and the result is very inaccurate after entering the place name"; extract the keywords "navigation function", "address search", "a bit inconvenient", and "the result is very inaccurate", then according to "very inconvenient" and "the result is very inaccurate", find the corresponding higher dissatisfaction emotion index A1 from the preset knowledge base; according to "navigation function", find the corresponding higher influence index B1 of the vehicle's normal driving from the preset knowledge base. Another example, when the usage comment is "The song name search of the music function is a bit inconvenient, and the result is a bit accurate after entering the song name"; extract the keywords "music function", "song name search", "the result is a bit inaccurate", and "a bit inconvenient", then according to "a bit inconvenient" and "the result is a bit inaccurate", find the corresponding lower dissatisfaction emotion index A2 from the preset knowledge base; according to "music function", find the corresponding lower influence index B2 of the vehicle's normal driving from the preset knowledge base.
[0053] Furthermore, according to the dissatisfaction sentiment index and the impact index on the normal driving of the vehicle, the first score of the function description text of the vehicle function is obtained. From the preset first mapping relation table, the sub-score corresponding to the dissatisfaction sentiment index is searched; the higher the dissatisfaction sentiment index, the lower the corresponding sub-score. From the preset second mapping relation table, the sub-score corresponding to the impact index on the normal driving of the vehicle is searched; the higher the impact index on the normal driving of the vehicle, the lower the corresponding sub-score. The sub-score corresponding to the dissatisfaction sentiment index and the sub-score corresponding to the impact index on the normal driving of the vehicle are weighted and averaged to obtain the first score of the function description text of the vehicle function.
[0054] For example, the sub-score corresponding to the dissatisfaction sentiment index is 6 points, and the corresponding weight coefficient is 40%; the sub-score of the impact index on the normal driving of the vehicle is 8 points, and the corresponding weight coefficient is 60%. Then, the first score of the function description text of the vehicle function is (40%×6 + 8×60%) = 7.2 points.
[0055] S203: Obtain the product documents of the target applications of multiple competing vehicles of the target model from the second network server 103.
[0056] Among them, the product documents of the target applications of each competing vehicle include the function description texts of multiple vehicle functions.
[0057] S204: For each vehicle function, compare and analyze the function description text of the vehicle function in the target application of the preset target model and the function description texts of the vehicle functions of each competing vehicle to obtain the second score of the function description text of the vehicle function.
[0058] Specifically, for each vehicle function, the function description text of the vehicle function in the target application of the preset target model and the function description texts of the vehicle functions of each competing vehicle are input into the pre-trained text comparison and analysis model to output the second score of the function description text of the vehicle function.
[0059] Among them, the text comparison and analysis model is obtained by training multiple second training samples input into the neural network to be trained. Each second training sample includes the historical function description texts of the historical vehicle functions of each historical competing vehicle of the historical model, the historical function description texts of the historical vehicle functions in the historical target application of the historical model, and the corresponding historical second scores.
[0060] S205: For each vehicle function, receive the third score of the function description text of the vehicle function in the target application of the preset target model from the expert terminal 104.
[0061] S206: For each vehicle function, perform a weighted average on the first score, second score, and third score of the functional description text of the vehicle function to obtain the comprehensive score of the functional description text of the vehicle function.
[0062] For example, if the first score is 6 points and the corresponding weight is 30%, the first score is 8 points and the corresponding weight is 20%; the third score is 5 points and the corresponding weight is 50%, then the comprehensive score of the functional description text of the vehicle function is equal to (30% × 6 + 8 × 20% + 50% × 5) = 5.9 points.
[0063] S207: Input the functional description texts corresponding to the top N vehicle functions with lower comprehensive scores into a pre-trained large model for optimizing text descriptions to obtain N optimized functional description texts.
[0064] Among them, the large model for optimizing text descriptions is obtained by training multiple first training samples input into the large model network to be trained. Each first training sample includes the functional description text before optimization in history and its corresponding historical comprehensive score, as well as the functional description text after optimization in history and its corresponding historical comprehensive score, where N is an integer greater than or equal to 1.
[0065] S208: Generate a new product document according to the N optimized functional description texts and a preset product document template.
[0066] Exemplarily, the product document may include a document number, a background introduction of the target application, the name of the target application, the names of each function of the target application and the corresponding functional description texts, etc., which are not limited herein.
[0067] In some embodiments, before S201, the method provided by the embodiments of the present application further includes:
[0068] Step 1: Obtain the operation behavior data of the vehicle functions of the target vehicle model from the vehicle supplier server. The operation behavior data includes at least the operation duration and the operation process steps.
[0069] Step 2: Determine the operation experience index of the vehicle functions of the target vehicle model according to the operation behavior data.
[0070] Step 3: Obtain the first score of the functional description text of the vehicle function according to the dissatisfaction emotion index, the influence index on the normal driving of the vehicle, and the operation experience index.
[0071] Exemplarily, the principle of Step 3 is the same as that of S202 described above, and will not be elaborated herein.
[0072] In some embodiments, the target vehicle model further includes: associated applications of the target application, where the associated applications are associated with function description texts corresponding to at least one of the top N vehicle functions with relatively low comprehensive scores. When the target application is a music application and the sound playback function is included in the top N vehicle functions with relatively low comprehensive scores, the associated application of the target application includes "radio", and the "radio" also includes the sound playback function.
[0073] The method provided by the embodiments of the present application further includes: synchronously updating the function description texts corresponding to at least one vehicle function associated with the associated application according to the optimized function description texts corresponding to at least one vehicle function. Exemplarily, the function description text of the sound playback function of the "radio" can be optimized according to the optimized function description text of the sound playback function of the music application.
[0074] In summary, for a product document generation method provided by the embodiments of the present application, for each vehicle function, a weighted average of the first score, the second score, and the third score of the function description text of the vehicle function can be obtained to get the comprehensive score of the function description text of the vehicle function.
[0075] The function description texts corresponding to the top N vehicle functions with relatively low comprehensive scores are input into a pre-trained large text description optimization model to obtain N optimized function description texts. Since the large text description optimization model is obtained by training a plurality of first training samples input into a large model network to be trained, and each first training sample includes the function description text before optimization and its corresponding historical comprehensive score in history, and the function description text after optimization and its corresponding historical comprehensive score in history. In this way, the function description texts corresponding to the top N vehicle functions with relatively low comprehensive scores are optimized. Furthermore, according to the N optimized function description texts and a preset product document template, a new product document is generated. In this way, the quality of the new product document is improved, enabling users and maintenance personnel to fully understand the new product document, and the efficiency of optimizing the product document is high, saving labor costs.
[0076] In addition, as Figure 3 shown, the embodiments of the present application provide a product document generation device. It should be noted that the basic principle and the technical effects generated by the product document generation device provided by the embodiments of the present application are the same as those of the above embodiments. For a brief description, for the parts not mentioned in the embodiments of the present application, reference can be made to the corresponding content in the above embodiments. The device provided by the embodiments of the present application includes a data acquisition unit, a text description scoring unit, a comprehensive score acquisition unit, a text description optimization unit, and a document generation unit, where
[0077] A data acquisition unit, configured to obtain usage comments of multiple vehicle functions integrated with a target application of a target vehicle model from a first network server 102.
[0078] A text description scoring unit, configured to, for each vehicle function, perform a quality assessment on the function description text of the vehicle function according to the corresponding usage comment, to obtain a first score of the function description text of the vehicle function.
[0079] The data acquisition unit is further configured to obtain product documents of the target application of multiple competing vehicle models of the target vehicle model from a second network server 103, wherein the product documents of the target application of each competing vehicle model include function description texts of multiple vehicle functions.
[0080] The data acquisition unit is further configured to, for each vehicle function, perform a comparison and analysis between the function description text of the vehicle function in the target application of the preset target vehicle model and the function description texts of the vehicle functions of each competing vehicle model, to obtain a second score of the function description text of the vehicle function.
[0081] The data acquisition unit is further configured to, for each vehicle function, receive a third score of the function description text of the vehicle function in the target application of the preset target vehicle model from an expert terminal 104.
[0082] A comprehensive score acquisition unit, configured to, for each vehicle function, perform a weighted average on the first score, the second score, and the third score of the function description text of the vehicle function, to obtain a comprehensive score of the function description text of the vehicle function.
[0083] A text description optimization unit, configured to input the function description texts corresponding to the top N vehicle functions with lower comprehensive scores into a pre-trained text description optimization large model, to obtain N optimized function description texts.
[0084] Wherein, the text description optimization large model is obtained by training multiple first training samples input into a large model network to be trained. Each first training sample includes the function description text before optimization in history and its corresponding historical comprehensive score, and the function description text after optimization in history and its corresponding historical comprehensive score, wherein N is an integer greater than or equal to 1.
[0085] A document generation unit, configured to generate a new product document according to the N optimized function description texts and a preset product document template.
[0086] In some embodiments, the text description scoring unit is specifically configured to, for each vehicle function, extract keywords associated with the usage problems of the vehicle function from the usage comments corresponding to the vehicle function; and perform a quality assessment on the function description text of the vehicle function according to the keywords associated with the usage problems to obtain a first score of the function description text of the vehicle function.
[0087] In some embodiments, the text description scoring unit is specifically configured to determine an unsatisfactory emotion index of the user for the usage problem and an influence index of the usage problem on the normal driving of the vehicle according to the keywords associated with the usage problems; and obtain a first score of the function description text of the vehicle function according to the unsatisfactory emotion index and the influence index on the normal driving of the vehicle.
[0088] In some embodiments, the text description scoring unit is specifically configured to look up the sub-score corresponding to the unsatisfactory emotion index from a preset first mapping relation table; look up the sub-score corresponding to the influence index on the normal driving of the vehicle from a preset second mapping relation table; and perform a weighted average on the sub-score corresponding to the unsatisfactory emotion index and the sub-score corresponding to the influence index on the normal driving of the vehicle to obtain a first score of the function description text of the vehicle function.
[0089] In some embodiments, the data acquisition unit is further configured to obtain operation behavior data of the user for the vehicle functions of the target model from the vehicle supplier server, where the operation behavior data at least includes the operation duration and the operation process steps. The text description scoring unit is specifically configured to determine an operation experience index of the user for the vehicle functions of the target model according to the operation behavior data; and obtain a first score of the function description text of the vehicle function according to the unsatisfactory emotion index, the influence index on the normal driving of the vehicle, and the operation experience index.
[0090] In some embodiments, the text description scoring unit is specifically configured to, for each vehicle function, input the function description text of the vehicle function in the target application of the preset target model and the function description texts of the vehicle functions of each competing vehicle into a pre-trained text comparison and analysis model to output a second score of the function description text of the vehicle function.
[0091] Wherein, the text comparison and analysis model is obtained by training a neural network to be trained with a plurality of second training samples, and each second training sample includes the historical function description texts of the historical vehicle functions of each historical competing vehicle of the historical model, the historical function description texts of the historical vehicle functions in the historical target application of the historical model, and the corresponding historical second scores.
[0092] In some embodiments, the target model further includes: an associated application of the target application, where the associated application is associated with the function description texts corresponding to at least one of the top N vehicle functions with relatively low comprehensive scores.
[0093] A synchronization update unit, configured to synchronously update the function description text corresponding to at least one vehicle function associated with an associated application according to the optimized function description text corresponding to at least one vehicle function.
[0094] In addition, an embodiment of the present application further provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the electronic device is caused to execute the method provided in the above embodiments of the present application.
[0095] In addition, an embodiment of the present application further provides a storage medium storing a computer program, which when executed by a processor, causes the computer to execute the method provided in the above embodiments of the present application.
[0096] In addition, an embodiment of the present application further provides a computer program product, including a computer program, which when run, causes an electronic device to execute the method provided in the above embodiments of the present application.
[0097] In the above description, no detailed description is made of the technical details such as the composition of each layer. However, those skilled in the art should understand that various technical means can be used to form layers, regions, etc. of the required shapes. In addition, in order to form the same structure, those skilled in the art can also design methods that are not exactly the same as the methods described above. In addition, although the above embodiments are described separately, this does not mean that the measures in each embodiment cannot be used advantageously in combination.
[0098] Although the preferred embodiments of the present application have been described, those skilled in the art can make additional changes and modifications once they learn the basic creative concept. Therefore, the appended claims are intended to be construed to include the preferred embodiments and all changes and modifications falling within the scope of the present application.
[0099] Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the claims of the present application and their equivalent technologies, the present application is also intended to include these changes and modifications.
Claims
1. A method for generating product documents, characterized in that: The method comprises: Obtaining user usage comments on multiple vehicle functions integrated in a target application of a target vehicle model from a first network server; for each of the vehicle functions, performing a quality assessment on a functional description text of the vehicle function according to the corresponding usage comments to obtain a first score for the functional description text of the vehicle function; Acquire, from a second network server, product documents of target applications of multiple competing vehicles of the target vehicle model, wherein the product document of the target application of each competing vehicle includes function description texts of the multiple vehicle functions; For each of the vehicle functions, a function description text of the vehicle function in the target application of the preset target vehicle model and a function description text of the vehicle function of each of the competing vehicles are compared and analyzed to obtain a second score of the function description text of the vehicle function; For each of the vehicle functions, receiving a third score of a function description text of the vehicle function in a target application of a preset target vehicle model from an expert terminal; For each of the vehicle functions, performing a weighted average of the first score, the second score, and the third score of the function description text of the vehicle function to obtain a comprehensive score of the function description text of the vehicle function; Inputting the function description texts corresponding to the first N vehicle functions with lower comprehensive scores into the pre-trained text description optimization large model to obtain N optimized function description texts, wherein the text description optimization large model is trained by inputting a plurality of first training samples into the large model network to be trained, each of the first training samples includes the function description text before historical optimization and its corresponding historical comprehensive score, and the function description text after historical optimization and its corresponding historical comprehensive score, wherein N is an integer greater than or equal to 1; Generate a new product document based on the optimized N function description texts and the preset product document template.
2. The method according to claim 1, characterized in that The step of performing a quality assessment on the functional description text of the vehicle function according to the corresponding usage comments to obtain a first score of the functional description text of the vehicle function includes: For each of the vehicle functions, extracting keywords associated with usage problems of the vehicle function from usage reviews corresponding to the vehicle function; A quality assessment is performed on the functional description text of the vehicle function according to the keywords associated with the usage question to obtain a first score of the functional description text of the vehicle function.
3. The method according to claim 2, characterized in that The step of performing a quality assessment on the functional description text of the vehicle function according to the keywords associated with the usage question to obtain a first score of the functional description text of the vehicle function includes: Determine, according to the keywords associated with the usage problem, a user's dissatisfaction sentiment index with respect to the usage problem and an impact index of the usage problem on normal driving of the vehicle; A first score of the function description text of the vehicle function is obtained according to the dissatisfaction sentiment index and the impact index on the normal driving of the vehicle.
4. The method according to claim 3, characterized in that The step of obtaining a first score of the function description text of the vehicle function according to the dissatisfaction sentiment index and the impact index on the normal driving of the vehicle includes: Searching for the sub-score corresponding to the dissatisfaction sentiment index from a preset first mapping relationship table; Searching for a sub-score corresponding to the influence index of normal driving of the vehicle from a preset second mapping relationship table; A weighted average is performed on the sub-score corresponding to the dissatisfaction sentiment index and the sub-score corresponding to the impact index on normal driving of the vehicle to obtain a first score of the function description text of the vehicle function.
5. The method according to claim 3, characterized in that: Before obtaining a first score of the function description text of the vehicle function according to the dissatisfaction sentiment index and the impact index on the normal driving of the vehicle, the method further includes: Acquire the user's operation behavior data on the vehicle function of the target vehicle model from the vehicle supplier server, wherein the operation behavior data at least includes the operation duration and the operation process steps; Determining, based on the operation behavior data, a user's operation experience index for a vehicle function of the target vehicle model; A first score of the function description text of the vehicle function is obtained according to the dissatisfaction emotion index, the impact index on the normal driving of the vehicle, and the operation experience index.
6. The method according to claim 1, characterized in that For each of the vehicle functions, comparing and analyzing the function description text of the vehicle function in the target application of the preset target vehicle model and the function description text of the vehicle function of each of the competing vehicles to obtain a second score of the function description text of the vehicle function, including: For each of the vehicle functions, the function description text of the vehicle function in the target application of the preset target vehicle model and the function description text of the vehicle function of each of the competing vehicles are input into a pre-trained text comparison and analysis model to output a second score for the function description text of the vehicle function, wherein the text comparison and analysis model is trained by inputting multiple second training samples into a neural network to be trained, and each of the second training samples includes historical function description texts of historical vehicle functions of each historical competing vehicle of the historical vehicle model, historical function description texts of historical vehicle functions in historical target applications of the historical vehicle model and corresponding historical second scores.
7. The method according to any one of claims 1 to 6, characterized in that: The target vehicle model further includes: an associated application of the target application, wherein the associated application is associated with a function description text corresponding to at least one vehicle function among the top N vehicle functions with lower comprehensive scores. After generating a new product document according to the optimized N function description texts and a preset product document template, the method further includes: According to the optimized function description text corresponding to the at least one vehicle function, the function description text corresponding to the at least one vehicle function associated with the associated application is synchronously updated.
8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the electronic device executes the method according to any one of claims 1 to 7.
9. A storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the computer is caused to perform the method according to any one of claims 1 to 7.
10. A computer program product, comprising a computer program, characterized in that When the computer program is executed, the electronic device is caused to execute the method as claimed in any one of claims 1 to 7.