Multi-language interface information generation method

Abstract

The invention provides a multi-language interface information generation method, which comprises the following steps of: acquiring first interface information based on a language interface switching instruction; the language interface switching instruction is used for instructing to convert first interface information of a first language into second interface information of a second language; determining a text type of first text information according to the first text information in the first interface information; determining text typesetting information of a second language according to the first text information, a text type of the first text information and character attributes of the second language; determining second text information of a second language according to the text type of the first text information, the text typesetting information of the second language, the first text information and the character attribute of the second language; generating second interface information of a second language according to first layout information and second text information in the first interface information; according to the scheme, the adaptation cost of the multi-language interface can be reduced, and the generation efficiency of the multi-language interface can be improved.

Description

Technical Field

[0001] This application relates to the field of computer technology, and in particular to a method for generating multilingual interface information. Background Technology

[0002] With the rapid development of the globalized economy, more and more products need to target the international market to adapt to the needs of users in different cultures, languages, and regions. For software products and internet services, implementing multilingual support and internationalized interfaces is a complex and time-consuming process.

[0003] Currently, traditional multilingual interfaces often rely on manual translation and proofreading to achieve adaptive display. This not only leads to high adaptation costs for multilingual interfaces but also results in low efficiency in generating multilingual interfaces. Summary of the Invention

[0004] This application provides a method for generating multilingual interface information, which can reduce the adaptation cost of multilingual interfaces and improve the generation efficiency of multilingual interfaces.

[0005] In a first aspect, embodiments of this application provide a method for generating multilingual interface information. This method can be executed by a multilingual interface information generation device, which can be a terminal device or a module for a terminal device, or a server or a module for a server. This application does not limit the executing entity of this method. The method includes: obtaining first interface information based on a language interface switching instruction; the language interface switching instruction is used to instruct the first interface information in the first language to be converted into second interface information in the second language; determining the text type of the first text information based on first text information in the first interface information; determining the text layout information of the second language based on the first text information, the text type of the first text information, and the character attributes of the second language; determining the second text information of the second language based on the text type of the first text information, the text layout information of the second language, the first text information, and the character attributes of the second language; and generating the second interface information of the second language based on the first layout information and the second text information in the first interface information.

[0006] The above solution, on the one hand, can automatically generate second interface information in a second language based on the first layout information and the second text information, without relying on manual translation and proofreading, thus reducing the adaptation cost of multilingual interfaces and improving the generation efficiency of multilingual interfaces; on the other hand, it determines the text layout information of the second language based on the text type of the first text information and the character attributes of the second language, which not only allows for different text layouts to be determined for different text types, but also automatically generates the text layout information of the second language. With the text layout information of the second language determined, the second interface information can be quickly generated based on the first layout information, thus improving the generation efficiency of multilingual interfaces.

[0007] In one possible implementation, a semantic importance score is determined based on the semantic encoding result of the first text information; a text type coefficient is determined based on the text type of the first text information; a fusion weight is determined based on the semantic importance score, the text type coefficient, the confidence level of the text layout information of the second language, and the confidence level precision coefficient; and a second text information of the second language is determined based on the fusion weight, the semantic encoding result of the first text information, the text layout information of the second language, and the character attributes of the second language.

[0008] The above scheme combines the semantic encoding results of the first text information with the text layout information of the second language in a weighted manner, which can solve the problem of insufficient feature representation of the existing single model and improve the accuracy of the second text information generation.

[0009] In one possible implementation, the text type is one of technical terminology, plain text, or explanatory text; the character attributes include one or more of the following: average length of each character, layout direction, and character density; the text layout information includes one or more of the following: text content per line, range of characters per line, line break probability, and number of lines.

[0010] Technical terms should prioritize semantic accuracy, general text should balance semantics and length, and explanatory text should prioritize avoiding length redundancy. The above solution can take into account the differences in layout requirements of text types in multilingual interfaces and improve the accuracy of text layout information generation.

[0011] In one possible implementation, second layout information in the second interface information is determined based on the first layout information; wherein the second layout information conforms to the rigid constraints of each component in the interface; and second interface information in the second language is generated based on the second layout information and the second text information.

[0012] The above solution clarifies the configurable scope and enforcement rules of each component, making the generated second interface information more accurate.

[0013] In one possible implementation, interface style information corresponding to the second language is determined; the interface style information is determined based on cultural feature data corresponding to the second language; the cultural feature data is used to indicate interface style information that conforms to the cultural habits and laws and regulations of the second language region; and second interface information in the second language is generated based on the interface style information corresponding to the second language, the second layout information, and the second text information.

[0014] The above solution can generate interface style information corresponding to the second language based on the cultural feature data corresponding to the second language.

[0015] In one possible implementation, the interface style information corresponding to the second language is determined from the cultural feature data corresponding to the second language based on the adaptation priority; the adaptation priority is determined based on one or more of the following: user-defined rules, compliance credibility, and user distribution ratio; the interface style information includes one or more of the following: font, color, warning color, disabled information, and icon information of each component in the second interface.

[0016] The above solution determines the adaptation priority based on user-defined rules, compliance credibility, and user distribution ratio, and then dynamically determines the interface style information corresponding to the second language based on the adaptation priority, which can increase the diversity of interface style information corresponding to the second language.

[0017] In one possible implementation, the interface style information corresponding to the second language, the second layout information, and the second text information are used to generate second interface information in the second language through a genetic algorithm and an objective function. The objective function is determined based on one or more of the following: overflow rate, truncation rate, readability, and path consistency. The overflow rate refers to the degree to which the text content of the second interface overflows within the second interface. The truncation rate refers to the degree to which technical terms are truncated. Readability refers to the degree to which the layout of the second interface is acceptable to the user. Path consistency refers to whether the function of any operable component in the second interface is consistent with that of the operable component corresponding to the first interface.

[0018] The above scheme can accurately evaluate the effectiveness of the second interface information through the objective function.

[0019] In one possible implementation, multiple first candidate interface information are generated based on the interface style information corresponding to the second language, the second layout information, the second text information, and a genetic algorithm; according to the objective function, a first optimal interface information is selected from the multiple first candidate interface information, and the first optimal interface information is determined as the second interface information of the second language.

[0020] The above scheme selects the first optimal interface information from multiple first candidate interface information, which can improve the effectiveness of the second interface information.

[0021] In one possible implementation, under the condition of satisfying iteration, the size, position, and interface style of any component in the first optimal interface information are adjusted to generate multiple second candidate interface information; the multiple second candidate interface information includes the first optimal interface information; according to the objective function, the optimal interface information selected from the multiple second candidate interface information is determined as the second interface information.

[0022] The above scheme adjusts the size, position, and interface style of any component in the first optimal interface information to generate multiple second candidate interface information, which can reduce the possibility of the second interface information getting trapped in a local optimum.

[0023] In one possible implementation, the second interface information is adjusted based on user feedback; a first target result of the adjusted second interface information in the objective function is determined; a second target result of the unadjusted second interface information in the objective function is determined; if the difference between the first target result and the second target result is greater than a first threshold, the user feedback information is determined to be invalid feedback.

[0024] The above solution removes invalid user feedback and adjusts the information on the second interface based on valid user feedback, thereby improving the effectiveness of the information on the second interface.

[0025] Secondly, embodiments of this application provide a multilingual interface information generation apparatus, comprising: an acquisition unit, a determination unit, and a generation unit. The acquisition unit is configured to acquire first interface information based on a language interface switching instruction; the language interface switching instruction is configured to instruct the first interface information in the first language to be converted into second interface information in the second language. The determination unit is configured to determine the text type of the first text information based on first text information in the first interface information; determine the text layout information of the second language based on the first text information, the text type of the first text information, and the character attributes of the second language; and determine the second text information of the second language based on the text type of the first text information, the text layout information of the second language, the first text information, and the character attributes of the second language. The generation unit is configured to generate the second interface information in the second language based on the first layout information in the first interface information and the second text information.

[0026] In one possible implementation, the determining unit is configured to: determine a semantic importance score based on the semantic encoding result of the first text information; determine a text type coefficient based on the text type of the first text information; determine a fusion weight based on the semantic importance score, the text type coefficient, the confidence level of the text layout information of the second language, and the confidence level precision coefficient; and determine the second text information of the second language based on the fusion weight, the semantic encoding result of the first text information, the text layout information of the second language, and the character attributes of the second language.

[0027] In one possible implementation, the text type is one of technical terminology, plain text, or explanatory text; the character attributes include one or more of the following: average length of each character, layout direction, and character density; the text layout information includes one or more of the following: text content per line, range of characters per line, line break probability, and number of lines.

[0028] In one possible implementation, the determining unit is configured to determine the second layout information in the second interface information based on the first layout information; wherein the second layout information conforms to the rigid constraints of each component in the interface; and the generating unit is configured to generate the second interface information in the second language based on the second layout information and the second text information.

[0029] In one possible implementation, the determining unit is used to determine interface style information corresponding to the second language; the interface style information is determined based on cultural feature data corresponding to the second language; the cultural feature data is used to indicate interface style information that conforms to the cultural habits and laws and regulations of the second language region; the generating unit is used to generate second interface information in the second language according to the interface style information corresponding to the second language, the second layout information, and the second text information.

[0030] In one possible implementation, the determining unit is used to determine the interface style information corresponding to the second language from the cultural feature data corresponding to the second language according to the adaptation priority; the adaptation priority is determined according to one or more of the following: user-defined rules, compliance credibility, user distribution ratio; the interface style information includes one or more of the following: font, color, warning color, disabled information, and icon information of each component in the second interface.

[0031] In one possible implementation, the generation unit is used to generate second interface information in the second language by using a genetic algorithm and an objective function to generate second interface information in the second language, based on the interface style information, the second layout information, and the second text information corresponding to the second language. The objective function is determined based on one or more of the following: overflow rate, truncation rate, readability, and path consistency. The overflow rate refers to the degree to which the text content of the second interface overflows in the second interface. The truncation rate refers to the degree to which technical terms are truncated. The readability refers to the degree to which the layout of the second interface is acceptable to the user. The path consistency refers to whether the function of any operable component in the second interface is consistent with the function of the operable component corresponding to the first interface.

[0032] In one possible implementation, the generation unit is configured to generate multiple first candidate interface information based on the interface style information corresponding to the second language, the second layout information, the second text information, and a genetic algorithm; select a first optimal interface information from the multiple first candidate interface information according to the objective function, and determine the first optimal interface information as the second interface information of the second language.

[0033] In one possible implementation, the generation unit is configured to adjust the size, position, and interface style of any component in the first optimal interface information under iterative conditions to generate a plurality of second candidate interface information; the plurality of second candidate interface information includes the first optimal interface information; the determining unit is configured to determine the optimal interface information selected from the plurality of second candidate interface information as the second interface information according to the objective function.

[0034] In one possible implementation, the determining unit is configured to: adjust the second interface information based on user feedback; determine a first target result of the adjusted second interface information in the objective function; determine a second target result of the unadjusted second interface information in the objective function; and determine that the user feedback is invalid if the difference between the first target result and the second target result is greater than a first threshold.

[0035] Thirdly, embodiments of this application also provide a computing device, including: Memory, used to store program instructions; The processor is configured to invoke program instructions stored in the memory and execute any method for implementing the first aspect described above, according to the obtained program instructions.

[0036] Fourthly, embodiments of this application also provide a computer-readable storage medium storing computer-readable instructions, which, when read and executed by a computer, implement any of the methods described in the first aspect.

[0037] Fifthly, embodiments of this application provide a computer program product, including a computer program executable by a computer device, which, when run on the computer device, causes the computer device to perform any of the methods described in the first aspect. Attached Figure Description

[0038] Figure 1 A flowchart illustrating a method for generating multilingual interface information provided in an embodiment of this application; Figure 2 A flowchart illustrating a method for determining second text information in a second language, provided as an embodiment of this application; Figure 3 A flowchart illustrating a method for generating multilingual interface information provided in an embodiment of this application; Figure 4 A schematic diagram of the structure of a multilingual interface information generation device provided in an embodiment of this application; Figure 5 This is a schematic diagram of a multilingual interface information generation device provided in an embodiment of this application. Detailed Implementation

[0039] This application provides a method for generating multilingual interface information, which can reduce the adaptation cost of multilingual interfaces and improve the generation efficiency of multilingual interfaces. The method, apparatus, and system are based on the same technical concept. Since the methods, apparatus, and systems solve problems in similar principles, their implementations can be mutually referenced, and repeated details will not be elaborated further.

[0040] The following explanations of some terms used in this application are provided to facilitate understanding by those skilled in the art.

[0041] (1) XLM-RoBERTa-Large, a pre-trained multilingual model used for semantic encoding and feature extraction of multilingual text.

[0042] (2) Long Short-Term Memory (LSTM) network, a deep learning model used for time series data prediction (such as text length prediction).

[0043] (3) Genetic Algorithm (GA) is a method that searches for the optimal solution by simulating the natural evolution process.

[0044] (4) Cascading Style Sheets (CSS) are used to describe the layout and style of UI elements on a webpage.

[0045] (5) Simulated annealing algorithm, which compares the objective function of the optimization problem to the energy of a physical system. The algorithm has a high probability of accepting poor solutions in the early stage (escaping local optima), and gradually reduces the probability of accepting poor solutions in the later stage, so as to stably approach the global optimum.

[0046] (6) The Kalman filter algorithm is an efficient recursive mathematical method used to estimate the unknown state of a dynamic system from a series of noisy observation data. Its core idea is to combine prediction (based on the model) and update (based on observation) to give the optimal estimate.

[0047] It should be understood that in the embodiments of this application, "at least one" means one or more, and "more than one" means two or more. "And / or" describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone, where A and B can be singular or plural. The character " / " generally indicates that the preceding and following related objects are in an "or" relationship. "At least one of the following" or similar expressions refer to any combination of these items, including any combination of single or plural items. For example, at least one of a, b, or c can represent: a, b, c, a and b, a and c, b and c, or a, b, and c, where a, b, and c can be single or multiple.

[0048] In the embodiments of this application, descriptions such as "when," "under the circumstances," "if," and "if" all refer to the electronic device making corresponding processing under certain objective circumstances. They are not time limits, nor do they require the electronic device to make a judgment action when implementing it, nor do they imply any other limitations.

[0049] It should be noted that in this application, "used for indicating" can include both direct and indirect indication. When describing information as being used to indicate A, it can include whether the information directly or indirectly indicates A, but does not necessarily mean that the information contains A. Taking first information used to indicate first content as an example, the first information can contain the first content, or a part of the first content, or an identifier, index, indicator, etc., of the first content, and can also contain algorithms, calculation parameters, etc., for determining the first content. This application does not limit the manner of "indication".

[0050] In this application, "containing / including A" may be equivalent to "containing / including information A". Information A is used to indicate A.

[0051] In addition, it should be understood that in the description of this application, the words "first" and "second" are used only for the purpose of distinguishing descriptions and should not be construed as indicating or implying relative importance or order.

[0052] With the rapid development of the globalized economy, more and more products need to target the international market to adapt to the needs of users in different cultures, languages, and regions. For software products and internet services, implementing multilingual support and internationalized interfaces is a complex and time-consuming process.

[0053] Currently, traditional multilingual interfaces often rely on manual translation and proofreading to achieve adaptive display. This not only leads to high adaptation costs for multilingual interfaces but also results in low efficiency in generating multilingual interfaces.

[0054] Based on this, embodiments of this application provide a method for generating multilingual interface information. This method can be executed by a device for generating multilingual interface information, which can be a terminal device or a module for a terminal device, or a server or a module for a server. This application does not limit the subject on which this method is executed. The method is as follows... Figure 1 As shown, it includes the following steps: Step 101: Obtain the first interface information based on the language interface switching instruction.

[0055] The language interface switching indicator is used to indicate the conversion of the first interface information in the first language to the second interface information in the second language.

[0056] Optionally, the user may issue a language interface switching instruction after clicking the language interface switching component on the first interface; or, the developers may generate the language interface switching instruction according to the user's needs; this application does not limit this.

[0057] Optionally, the language interface switching indicator includes first interface information, a first language identifier, and a second language identifier. The first interface information includes first text information and first layout information; the first text information includes the text content of each component within the first interface; the first layout information includes the size, position, color, etc., of each component within the first interface. For example, the first interface information may include a UI component tree (carrying attributes such as component ID, type, text content, priority, etc.), a second language identifier (e.g., "DE_DE"), and device parameters (screen resolution, device type, ambient lighting).

[0058] Optionally, required fields (such as component tree structures) can be checked according to a structured data validation mechanism. If a core field is missing, a standardized error flag is generated and reported.

[0059] Optionally, the information on the first interface can also be normalized. For example, the device parameter normalization algorithm can be called to convert the physical screen resolution into the logical pixel matrix dimension. At the same time, the ambient light parameters can be mapped to preset enumeration values ​​(for example, "outdoor strong light" can be mapped to a specific environment identifier, and the overall color of the interface should be brighter under strong light conditions), and the normalized parameters can be stored in a temporary cache.

[0060] Step 102: Determine the text type of the first text information based on the first text information in the first interface information.

[0061] Optionally, the text type can be one of technical terminology, general text, or explanatory text. Other types are also possible, and this application does not limit this. Technical terminology text refers to precisely defined text content used in a specific field (such as medicine, law, engineering, finance, computer science, etc.); general text refers to everyday vocabulary, which is the primary text content for daily communication, information acquisition, and entertainment for the general public; explanatory text refers to text content whose primary goal is to explain, illustrate, or guide, elucidating how, why, or how something works.

[0062] This application does not limit the method for determining the text type of the first text information based on the first text information in the first interface information. For example, the first text information can be input into a neural network model, and the text type of the first text information can be determined based on the output of the neural network model. The neural network model can be the XLM-RoBERTa-Large model, or other models, and this application does not limit it.

[0063] Step 103: Determine the text layout information of the second language based on the first text information, the text type of the first text information, and the character attributes of the second language.

[0064] Optionally, character attributes include one or more of the following: average length per character, layout direction, and character density. The average length per character can be determined based on the total length of all characters in the second language and the number of characters in the second language. Average length per character typically refers to the horizontal space occupied by a character in a monospaced font (such as a code font); layout direction refers to the output order of characters, for example, left alignment in German; character density can be understood as the amount of semantic information contained within a unit area (such as a line or a page).

[0065] Of course, character attributes can include other attributes, but this application does not limit this.

[0066] Optionally, the text formatting information includes one or more of the following: the text content of each line, the character count range of each line, the line break probability, and the number of lines. For example, if there are 2 lines, the text content of the first line is XXXXX, the character count of the first line is 10-20, and the line break probability of the first line is 90%; the text content of the second line is YYYYYY, the character count of the first line is 15-20, and the line break probability of the first line is 10%.

[0067] Of course, text formatting information may include other content, and this application does not limit this.

[0068] This application does not limit the specific method for determining the text layout information of the second language based on the first text information, the text type of the first text information, and the character attributes of the second language. For example, the first text information, the text type of the first text information, and the character attributes of the second language can be input into a neural network model, and the text layout information of the second language can be determined based on the output of the neural network model. For example, the neural network module can be an LSTM model, but it can also be other models, and this application does not limit it.

[0069] Step 104: Determine the second text information in the second language based on the text type of the first text information, the text layout information of the second language, the first text information, and the character attributes of the second language.

[0070] Step 105: Generate second interface information in the second language based on the first layout information and the second text information.

[0071] Optionally, the second interface information in the second language can be rendered, or the second interface information in the second language can be stored and then rendered by the front-end device. For example, cultural feature parameters can be combined to generate dynamic layout rules that conform to the front-end rendering standard and stored in the content delivery network (CDN) to generate a unique access identifier.

[0072] Optionally, the terminal device compiles the dynamic layout rules into binary instructions via WebAssembly and calls the browser's hardware acceleration interface to render the interface. The system monitors the entire rendering process in real time. If it exceeds a 200-millisecond threshold, it automatically enables simplified rendering strategies (such as disabling non-core visual effects) to ensure responsiveness. WebAssembly (Wasm) is a high-performance, portable binary instruction format designed for the Web. Traditional Web applications rely entirely on JavaScript, but JavaScript's performance and efficiency are bottlenecks when handling computationally intensive tasks (such as games, image / video editing, 3D rendering, scientific computing, and encryption). Wasm aims to be the "high-performance computing engine" for the Web.

[0073] Optionally, a reusable template can be generated and stored in a template library based on the text type of the first text information, the text layout information of the second language, the character attributes of the first text information and the second language, and the second text information of the second language. Alternatively, structured indicator data and template information can be input into a report generation tool to automatically generate a standardized adaptation report containing indicator charts, parameter details, and compliance explanations for developers to trace and evaluate.

[0074] Optionally, the element tree of the rendered second interface is traversed to calculate the text overflow rate; based on the Web Content Accessibility Guidelines, outdoor readability scores are calculated by combining parameters such as font contrast and font size, and the adaptation effect indicators of the second interface are quantified.

[0075] The above solution, on the one hand, can automatically generate second interface information in a second language based on the first layout information and the second text information, without relying on manual translation and proofreading, thus reducing the adaptation cost of multilingual interfaces and improving the generation efficiency of multilingual interfaces; on the other hand, it determines the text layout information of the second language based on the text type of the first text information and the character attributes of the second language, which not only allows for different text layouts to be determined for different text types, but also automatically generates the text layout information of the second language. With the text layout information of the second language determined, the second interface information can be quickly generated based on the first layout information, thus improving the generation efficiency of multilingual interfaces.

[0076] In one possible implementation, step 104 above, which determines the second text information of the second language based on the text type of the first text information, the text layout information of the second language, the first text information, and the character attributes of the second language, is as follows: Figure 2 As shown, it includes the following steps: Step 201: Determine the semantic importance score based on the semantic encoding result of the first text information.

[0077] The text type is obtained based on the semantic encoding result of the first text information, where the semantic importance score can be the confidence level of each text type. For example, word segmentation and stop word removal are performed on the first text information (e.g., processing "fault code: overload" into a sequence), and then the preprocessed text information is loaded into a pre-fine-tuned XLM-RoBERTa-Large neural network model to obtain a high-dimensional semantic vector. The probability distribution of the text type is calculated through a fully connected layer to label the text type (divided into: technical terms, ordinary text, and explanatory text). The maximum confidence level is then determined as the semantic importance score. For example, if the confidence level of technical terms output by the neural network model is 80%, the confidence level of ordinary text is 10%, and the confidence level of explanatory text is 10%, and the confidence level of technical terms is the highest, then the text type of the first text information is determined to be technical terms, and the semantic importance score is 80%.

[0078] Step 202: Determine the text type coefficient based on the text type of the first text information.

[0079] The text type coefficient can be obtained by training a neural network model or by historical experience. This application does not limit this. For example, the text type is obtained based on the semantic encoding result of the first text information. The text type coefficient can be set to 1.3 (technical terminology), 1.0 (ordinary text), and 0.8 (explanatory text) according to the text category.

[0080] Step 203: Determine the fusion weights based on the semantic importance score, text type coefficient, confidence level of the second language text layout information, and confidence level precision coefficient.

[0081] Optionally, the confidence precision coefficient can be obtained from training the neural network model or from historical experience. This application does not limit this. For example, the confidence precision coefficient can be fixed at 1.0 to simplify the calculation and reserve space for multi-model expansion.

[0082] Optionally, the first text information, the text type of the first text information, and the character attribute time of the second language are input into the LSTM neural network model to obtain the confidence level of the second language typesetting information.

[0083] Optionally, if the confidence level of the second language typesetting information is less than the first threshold, the neural network model can be adjusted, or the input can be fed into another neural network. This application does not limit the first threshold. For example, if the first threshold is 95%, and the model outputs a 95% confidence interval for the second language typesetting information, it can ensure that the mean absolute error (MAE) is controlled within 5%.

[0084] Optionally, the method for determining the fusion weights is shown in the following formula (1): Fusion weight = semantic importance score × text type coefficient + confidence score × confidence precision coefficient (1) Step 204: Based on the fusion weight, the semantic encoding result of the first text information, the text layout information of the second language, and the character attributes of the second language, determine the second text information of the second language.

[0085] Optionally, the semantic encoding result of the first text information and the text layout information of the second language are multiplied by the fusion weight to obtain the fusion result, and the fusion result and the character attributes of the second language are concatenated to form the second text information of the second language. Of course, the second text information of the second language can also be determined by other methods, and this application does not limit this.

[0086] This application addresses the insufficient representation of semantic encoding features and text layout features in existing technologies, which exist independently. Therefore, it incorporates the core evaluation indicators of both types of features to ensure that the fusion result balances semantic relevance with layout adaptation accuracy. Considering the different layout requirements of text types (technical terms, general text, and explanatory text) in multilingual interfaces—technical terms require priority in ensuring semantic accuracy, general text requires a balance between semantics and length, and explanatory text requires priority in avoiding length redundancy—a text type coefficient is specifically introduced to achieve dynamic weight adjustment, rather than using a fixed weight design. Simultaneously, the confidence level is predicted through an association model (calculated inversely from the MAE error of the LSTM model; when the error is ≤5%, the confidence level is ≥0.95), ensuring that high-confidence features receive higher weights and further improving the reliability of the fusion result. The key parameter values ​​are all calibrated based on actual scenarios: the semantic importance score comes from the semantic encoding result of the neural network model; the text type coefficient is set to 1.3 (technical terms), 1.0 (general text), and 0.8 (explanatory text) according to text category; the precision coefficient is fixed at 1.0 to simplify calculation and reserve space for multi-model expansion.

[0087] This application provides a second implementation method, which differs from steps 101 to 105 described above in that it can further determine the second layout information in the second interface information based on the first layout information; then, based on the second layout information and the second text information, it generates second interface information in a second language. The second layout information conforms to the rigid constraints of each component in the interface; for example, configuration rules are parsed to extract the rigid constraints of the components (e.g., "fault indication area cannot shrink" and "operation button minimum width ratio 15%"). The percentage-based size constraints are converted into specific pixel value ranges (e.g., [205px, 300px]) based on the current pixel matrix, generating a standardized constraint dictionary that clarifies the configurable range and mandatory rules for each component, serving as boundary conditions for subsequent layout optimization.

[0088] This application provides a third embodiment, which differs from steps 101 to 105 and the second embodiment described above in that it further determines interface style information corresponding to the second language; then, based on the interface style information, second layout information, and second text information corresponding to the second language, it generates second interface information for the second language. The interface style information is determined based on cultural feature data corresponding to the second language; the cultural feature data indicates interface style information that conforms to the cultural customs and laws and regulations of the second language region; the interface style information includes one or more of the following for each component in the second interface: font, color, warning color, disabled information, and icon information.

[0089] For example, based on the language identifier of the second language, a knowledge graph query engine is launched to obtain the cultural feature data corresponding to the second language. Specifically, using "target language (German) + region (Germany) + application scenario (charging pile)" as the combined query conditions, the corresponding cultural feature data is retrieved from the knowledge graph query engine, including recommended fonts, warning colors, and operation icon specifications.

[0090] Optionally, this application can also determine the interface style information corresponding to the second language from the cultural feature data corresponding to the second language based on the adaptation priority; the adaptation priority is determined based on one or more of the following: user-defined rules, compliance credibility, and user distribution ratio. Users can customize the font, color, icon, etc. of each component corresponding to the second language, as well as the interface style information in different scenarios; compliance credibility refers to the quantitative verification results from the official standard API, such as the quantitative verification results from the official VDE standard API, where VDE is the German Association for Electrical, Electronic & Information Technology; user distribution ratio refers to the distribution of users in different countries and regions.

[0091] Optionally, user-defined rules are used as the primary basis, with compliance credibility and user distribution ratio as secondary factors. For example, if the user-defined color of the first component in the second interface is red, then the color of the first component is set to red. If there is no rule for the color of the first component in the user-defined rules, then the color of the first component can be determined based on compliance credibility and user distribution ratio. For example, if the usage rate of German personnel in the second interface is relatively high, and German personnel tend to prefer yellow, and yellow also meets the compliance credibility requirements, then the color of the first component is set to yellow.

[0092] Optionally, the adaptation priority is dynamically determined based on the weights of user-defined rules, compliance credibility, and user distribution ratio. This application does not limit the specific values ​​of the weights for user-defined rules, compliance credibility, and user distribution ratio. For example, if the first component in the user-defined rules is red, the VDE standard match between red and yellow is 0.98, and the proportion of German users is 0.8; the weights of the user-defined rules, compliance credibility, and user distribution ratio are 0.6, 0.2, and 0.2 respectively; then the probability that the first component is red is 0.6. 1+0.2 0.98 + 0.2 0.2 = 0.836; the probability that the first component is yellow is 0.6. 0+0.2 0.98 + 0.2 0.8 = 0.356; therefore, the color of the first component can be set to red.

[0093] Optionally, a new algorithm for resolving cultural rule conflicts can be added. When cultural characteristics of multiple regions conflict (such as the same icon being compliant in region A but prohibited in region B), the algorithm will automatically call a third-party compliance database for verification and intelligently decide the optimal solution based on the distribution ratio of user groups (e.g., if users in region B account for more than 60%, the rules of region B will be given priority).

[0094] In one implementation, the interface style information, layout information, and text information corresponding to the second language can be used to generate second interface information in the second language through a genetic algorithm and an objective function. The method is as follows: Figure 3 As shown, it includes the following steps: Step 301: Generate multiple first candidate interface information based on the interface style information corresponding to the second language, the second layout information, the second text information, and the genetic algorithm.

[0095] Optionally, the interface style information, second layout information, and second text information corresponding to the second language are used as input, and multiple first candidate interface information are randomly generated according to the size of the initial population. For example, a hybrid mode of "real number encoding + binary encoding" is adopted: continuous parameters such as component width and height are encoded with real numbers (e.g., 180.0px), and the initial range of parameters comes from low-code constraint rules (e.g., the minimum width of the button occupies 15% of the screen), the current device resolution mapping (e.g., 1920×1080 screen conversion), and the 95% confidence interval of historical adaptation data of similar components; discrete parameters such as layout direction and whether to wrap are encoded with binary (01 represents left alignment); the initial population size is 100, and multiple first candidate interface information are randomly generated (e.g., the width range of the first component [205, 300], the height range [40, 60]).

[0096] Step 302: Based on the objective function, select the first optimal interface information from multiple first candidate interface information, and determine the first optimal interface information as the second interface information of the second language.

[0097] The objective function is determined based on one or more of the following: overflow rate, truncation rate, readability, and path consistency. Overflow rate refers to the extent to which the text content of the second interface overflows in the second interface. Truncation rate refers to the extent to which technical terms are truncated. Readability refers to the degree to which the layout of the second interface is acceptable to the user. Path consistency refers to whether the function of any operable component in the second interface is consistent with the corresponding operable component in the first interface.

[0098] The objective function can be represented by the following formula (2): Objective function = A1×(1-overflow rate) + A2×(1-truncation rate) + A3×readability + A4×path consistency (2) Among them, A1, A2, A3 and A4 are weight parameters, and this application does not limit their specific values.

[0099] Optionally, the overflow rate is obtained by calculating and normalizing the total pixel difference between the content requirement size of the component (from the version information prediction result in step 103) and the current allocated size; the truncation rate is specifically calculated for key text components marked as "technical terms" to measure the proportion of information omitted or cut off due to insufficient allocated size; the readability index can be calculated based on second language specifications (such as German font size 14px, contrast ratio 3:1, which can be set manually based on experience); path consistency is calculated based on the difference in click steps between the first interface and the second interface. For example, in the search function, if the search is completed by entering search information and clicking the search button in the first interface, and the search can also be completed in the second interface through the above two steps, it indicates that the search component in the second interface has the same function as the search component in the first interface. The closer the objective function value is to 1, the better the information in the second interface.

[0100] Alternatively, since the genetic algorithm (GA) may get stuck in local optima, simulated annealing (SA) can be used as an auxiliary search.

[0101] Optionally, under the iterative condition, the size, position, and interface style of any component in the first optimal interface information are adjusted to generate multiple second candidate interface information; the multiple second candidate interface information includes the first optimal interface information; according to the objective function, the optimal interface information selected from the multiple second candidate interface information is determined as the second interface information.

[0102] Specifically, the optimal solution in the GA population, i.e., the first optimal interface information. This serves as the initial state of SA.

[0103] Under the condition of satisfying iteration, by... Make random, minor adjustments to generate new candidate solutions, which are also multiple second candidate interface information. .

[0104] Calculate the fitness difference based on the objective function. = - .like If the value is >0 (the new candidate solution is better), then it is accepted unconditionally. If the value is less than 0 (the new candidate solution is worse), then it can be decided whether to accept the worse solution.

[0105] When the loop ends, the final solution found by SA is injected back into the GA population. By accepting worse solutions, SA successfully helps GA escape local optima, thus ensuring its global optimization capability.

[0106] In one possible implementation, the information on the second interface can also be adjusted based on user feedback. For example, when a user clicks the "Layout Feedback" button, the front end collects the feedback content, component ID, and current device parameters, and uploads them to the database in JSON format, forming a complete feedback data chain.

[0107] However, user feedback may be invalid, so invalid user feedback can be removed; for example, determine the first target result of the adjusted second interface information in the objective function; determine the second target result of the unadjusted second interface information in the objective function; if the difference between the first target result and the second target result is greater than a first threshold, then determine that the user feedback information is invalid feedback.

[0108] Alternatively, the Kalman filter algorithm can be used to determine whether the user's feedback is invalid.

[0109] The specific execution steps of the Kalman filter algorithm are as follows: State initialization: At the start of the adaptation task, initialize the initial estimate of the fault confidence level of the target parameters (such as the line spacing coefficient). and initial covariance .

[0110] State prediction: based on the optimal state at the previous time step. Predict the prior state at time k using the state transition matrix A. :

[0111] Kalman gain calculation: Receive the raw feedback signal uploaded by the user. This was used as a measurement. The Kalman gain was then calculated. This gain determines the effect on new measurements. The degree of trust determines whether the user's feedback is invalid. If the degree of trust is low, the user's feedback is determined to be invalid; if the degree of trust is high, the user's feedback is determined to be valid.

[0112] State Update: Utilizing Kalman Gain Fusion Prediction Status and measured values The current optimal state is calculated. (Second interface information adjusted based on user feedback):

[0113] (This step completes the processing of the original feedback) (Noise reduction processing)

[0114] KF filters through time-series tracking and prediction-correction mechanisms. The system eliminates erroneous noise to ensure that only genuine, reliable negative signals enter the learning model. The reinforcement learning model marks this effective feedback as a negative reward and automatically adjusts the algorithm parameter configuration file to achieve closed-loop self-evolution of the system.

[0115] This application addresses the issue of insufficient feature representation in existing single-model approaches by designing a cross-modal feature fusion module. This module weights and fuses the semantic features (such as text type and technical terminology tags) output by XLM-RoBERTa-Large with the version information prediction features (such as character count range and line break probability) from LSTM. The fusion weights are dynamically adjusted through an attention mechanism (e.g., the semantic feature weights of technical terminology texts are increased by 30%). The input layer structure of the LSTM model is optimized by introducing prior features such as character density and layout direction of the target language, reducing the text length prediction error rate from over 15% in existing technologies to below 5%.

[0116] To address the static nature of traditional knowledge graphs, a dynamic weighted knowledge graph is constructed. Dynamic attributes such as scene priority, regional credibility, and time validity are set for each cultural feature (e.g., font, color). Feature adaptation priority is calculated in real time using the weighted formula (2).

[0117] A new algorithm for resolving cultural rule conflicts has been added. When cultural characteristics conflict between multiple regions (such as the same icon being compliant in region A but prohibited in region B), the algorithm automatically calls a third-party compliance database for verification and intelligently decides the optimal solution based on the distribution ratio of user groups (e.g., if users in region B account for more than 60%, the rules of region B will be used first).

[0118] An adaptive genetic-simulated annealing hybrid algorithm is proposed, which adopts a hybrid mode of "real number encoding + binary encoding" to adapt to continuous / discrete parameters; the crossover mutation probability is adaptively adjusted according to the population fitness variance (increased by 20% when variance > 0.3, and decreased by 15% when variance < 0.1); the Metropolis criterion of the simulated annealing algorithm is introduced, which allows for the acceptance of poor solutions when the iteration gets stuck in local optima, avoids algorithm stagnation, improves the convergence speed of layout optimization by 40%, and keeps the text overflow rate and element overlap rate at a low level.

[0119] A hierarchical feedback denoising model was constructed, simultaneously collecting explicit feedback (user-initiated annotations) and implicit feedback (such as user dwell time when text overflows and coordinates of accidental clicks). Invalid feedback (such as misoperation annotations) was filtered out using a Kalman filter algorithm, achieving a denoising rate of 85%. A "feedback-parameter adjustment" rule base was established, directly mapping feedback data to algorithm parameters (such as increasing the line spacing coefficient by 0.1 for "text too dense"), enabling real-time automatic iteration of model parameters, replacing manual adjustment, and improving iteration efficiency by 60%.

[0120] A new intelligent template generation module has been added. Based on a cultural feature knowledge base and historical adaptation cases, it automatically generates multilingual layout templates for different industries (such as finance and e-commerce) through transfer learning, and supports one-click reuse.

[0121] It integrates real-time preview simulation function and uses computer vision technology to simulate the interface display effect under different devices (such as outdoor touch screens and mobile phones) and different environments (such as strong light and low light), detects layout problems in advance (such as insufficient font recognition under strong light), and reduces the cost of later debugging.

[0122] Based on the same technological concept Figure 4 An exemplary illustration shows a multilingual interface information generation apparatus 400 provided in an embodiment of this application. For example... Figure 4As shown, it includes: an acquisition unit 401, a determination unit 402, and a generation unit 403. The acquisition unit 401 is used to acquire first interface information based on a language interface switching instruction; the language interface switching instruction is used to instruct the first interface information in the first language to be converted into second interface information in the second language. The determination unit 402 is used to determine the text type of the first text information based on first text information in the first interface information; determine the text layout information of the second language based on the first text information, the text type of the first text information, and the character attributes of the second language; and determine the second text information of the second language based on the text type of the first text information, the text layout information of the second language, the first text information, and the character attributes of the second language. The generation unit 403 is used to generate the second interface information in the second language based on the first layout information in the first interface information and the second text information.

[0123] In one possible implementation, the determining unit 402 is configured to: determine a semantic importance score based on the semantic encoding result of the first text information; determine a text type coefficient based on the text type of the first text information; determine a fusion weight based on the semantic importance score, the text type coefficient, the confidence level of the text layout information of the second language, and the confidence level precision coefficient; and determine the second text information of the second language based on the fusion weight, the semantic encoding result of the first text information, the text layout information of the second language, and the character attributes of the second language.

[0124] In one possible implementation, the text type is one of technical terminology, plain text, or explanatory text; the character attributes include one or more of the following: average length of each character, layout direction, and character density; the text layout information includes one or more of the following: text content per line, range of characters per line, line break probability, and number of lines.

[0125] In one possible implementation, the determining unit 402 is used to determine the second layout information in the second interface information based on the first layout information; wherein the second layout information conforms to the rigid constraint conditions of each component in the interface; and the generating unit 403 is used to generate the second interface information in the second language based on the second layout information and the second text information.

[0126] In one possible implementation, the determining unit 402 is used to determine the interface style information corresponding to the second language; the interface style information is determined based on the cultural feature data corresponding to the second language; the cultural feature data is used to indicate the interface style information that conforms to the cultural habits and laws and regulations of the second language region; the generating unit 403 is used to generate the second interface information of the second language according to the interface style information corresponding to the second language, the second layout information, and the second text information.

[0127] In one possible implementation, the determining unit 402 is used to determine the interface style information corresponding to the second language from the cultural feature data corresponding to the second language according to the adaptation priority; the adaptation priority is determined according to one or more of the following: user-defined rules, compliance credibility, user distribution ratio; the interface style information includes one or more of the following: font, color, warning color, disabled information, icon information of each component in the second interface.

[0128] In one possible implementation, the generation unit 403 is used to generate second interface information in the second language by using the interface style information corresponding to the second language, the second layout information, and the second text information through a genetic algorithm and an objective function; wherein, the objective function is determined based on one or more of the following: overflow rate, truncation rate, readability, and path consistency; the overflow rate refers to the degree to which the text content of the second interface overflows in the second interface; the truncation rate refers to the degree to which technical terms are truncated; the readability refers to the degree to which the layout of the second interface is acceptable to the user; and the path consistency refers to whether the function of any operable component in the second interface is consistent with that of the operable component corresponding to the first interface.

[0129] In one possible implementation, the generation unit 403 is configured to generate a plurality of first candidate interface information based on the interface style information corresponding to the second language, the second layout information, the second text information, and a genetic algorithm; select a first optimal interface information from the plurality of first candidate interface information according to the objective function, and determine the first optimal interface information as the second interface information of the second language.

[0130] In one possible implementation, the generation unit 403 is used to adjust the size, position, and interface style of any component in the first optimal interface information under the condition of satisfying iteration, and generate a plurality of second candidate interface information; the plurality of second candidate interface information includes the first optimal interface information; the determining unit 402 is used to determine the optimal interface information selected from the plurality of second candidate interface information as the second interface information according to the objective function.

[0131] In one possible implementation, the determining unit 402 is configured to adjust the second interface information based on user feedback; determine the first target result of the adjusted second interface information in the target function; determine the second target result of the unadjusted second interface information in the target function; and if the difference between the first target result and the second target result is greater than a first threshold, then determine that the user feedback is invalid feedback.

[0132] Based on the same technical concept, embodiments of this application provide a multilingual interface information generation device 500, which may be, for example, a computing device. Figure 5 As shown, a multilingual interface information generation apparatus 500 includes at least one processor 501 and a memory 502 connected to the at least one processor. In this embodiment, the specific connection medium between the processor 501 and the memory 502 is not limited. Figure 5 Taking the connection between processor 501 and memory 502 via a bus as an example, the bus can be divided into address bus, data bus, control bus, etc.

[0133] In this embodiment of the application, the memory 502 stores instructions that can be executed by at least one processor 501. By executing the instructions stored in the memory 502, at least one processor 501 can execute the above-described method for generating multilingual interface information.

[0134] The processor 501 serves as the control center of a multilingual interface information generation device 500. It can connect to various parts of a computer device via various interfaces and lines, and performs resource settings by running or executing instructions stored in the memory 502 and calling data stored in the memory 502. Optionally, the processor 501 may include one or more determining units. The processor 501 may integrate an application processor and a modem processor, wherein the application processor mainly handles the operating system, user interface, and applications, while the modem processor mainly handles wireless communication. It is understood that the modem processor may not be integrated into the processor 501. In some embodiments, the processor 501 and the memory 502 may be implemented on the same chip; in some embodiments, they may also be implemented on separate chips.

[0135] Processor 501 can be a general-purpose processor, such as a central processing unit (CPU), digital signal processor, application-specific integrated circuit (ASIC), field-programmable gate array (FPGA), or other programmable logic device, discrete gate or transistor logic device, or discrete hardware component, capable of implementing or executing the methods, steps, and logic block diagrams disclosed in the embodiments of this application. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the methods disclosed in the embodiments of this application can be directly manifested as being executed by a hardware processor, or executed by a combination of hardware and software modules within the processor.

[0136] Memory 502, as a non-volatile computer-readable storage medium, can be used to store non-volatile software programs, non-volatile computer-executable programs, and modules. Memory 502 may include at least one type of storage medium, such as flash memory, hard disk, multimedia card, card-type memory, random access memory (RAM), static random access memory (SRAM), programmable read-only memory (PROM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), magnetic memory, magnetic disk, optical disk, etc. Memory 502 can be any other medium capable of carrying or storing desired program code in the form of instructions or data structures that can be accessed by a computer, but is not limited thereto. In the embodiments of this application, memory 502 can also be a circuit or any other device capable of implementing storage functions for storing program instructions and / or data.

[0137] This application also provides a computer-readable storage medium storing a computer-executable program, which is used to cause a computer to execute a method for generating multilingual interface information as listed in any of the above methods.

[0138] This application provides a computer program product, including a computer program executable by a computer device. When the program is run on the computer device, it causes the computer device to execute a method for generating multilingual interface information as listed in any of the above methods.

[0139] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0140] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to this application. It should be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0141] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0142] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0143] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the spirit and scope of this application. Therefore, if such modifications and variations fall within the scope of the claims of this application and their equivalents, this application also intends to include such modifications and variations.

Claims

1. A method of generating multi-lingual interface information, characterized by, The method comprises: obtaining first interface information based on a language interface switching instruction; the language interface switching instruction is used to indicate conversion of first interface information in a first language into second interface information in a second language; determining a text type of first text information in the first interface information according to the first text information; determining text layout information of the second language according to the first text information, the text type of the first text information, and character attributes of the second language; determining second text information in the second language according to the text type of the first text information, the text layout information of the second language, the first text information, and the character attributes of the second language; generating second interface information in the second language according to first layout information in the first interface information and the second text information.

2. The method of claim 1, wherein, The determination of the second text information in the second language according to the text type of the first text information, the text layout information of the second language, the first text information, and the character attributes of the second language comprises: determining a semantic importance score based on a semantic coding result of the first text information; determining a text type coefficient based on a text type of the first text information; determining a fusion weight according to the semantic importance score, the text type coefficient, a confidence degree of the text layout information of the second language, and a confidence degree precision coefficient; determining the second text information in the second language based on the fusion weight, the semantic coding result of the first text information, the text layout information of the second language, and the character attributes of the second language.

3. The method of claim 1, wherein, The text type is one of a professional term, a general text, or an explanatory text. The character attributes include one or more of an average length of each character, a layout direction, and a character density. The text layout information includes one or more of text content of each line, a character number interval of each line, a line break probability, and a line number.

4. The method according to any one of claims 1 to 3, characterized in that, The method further comprises: determining second layout information in the second interface information according to the first layout information; wherein the second layout information meets a rigidity constraint condition of each component in an interface; The generation of the second interface information in the second language according to the first layout information in the first interface information and the second text information comprises: generating the second interface information in the second language according to the second layout information and the second text information.

5. The method of claim 4, wherein, The method further comprises: determining interface style information corresponding to the second language; the interface style information is determined based on cultural characteristic data corresponding to the second language; the cultural characteristic data is used to indicate interface style information meeting cultural habits and laws and regulations in a region of the second language; The generation of the second interface information in the second language according to the second layout information and the second text information comprises: generating the second interface information in the second language according to the interface style information corresponding to the second language, the second layout information, and the second text information.

6. The method of claim 5, wherein, The determination of the interface style information based on the cultural characteristic data corresponding to the second language comprises: determine interface style information corresponding to the second language from cultural feature data corresponding to the second language according to an adaptation priority; the adaptation priority is determined according to one or more of the following: a rule defined by the user, compliance credibility, and user distribution ratio; the interface style information includes one or more of the following: font, color, warning color, disabled information, and icon information of each component in the second interface.

7. The method of claim 5, wherein, generating the second interface information in the second language according to the interface style information corresponding to the second language, the second layout information, and the second text information includes: generating the second interface information in the second language by using a genetic algorithm and an objective function according to the interface style information corresponding to the second language, the second layout information, and the second text information; the objective function is determined according to one or more of the following: overflow rate, truncation rate, readability, and path consistency; the overflow rate refers to the overflow degree of text content in the second interface in the second interface; the truncation rate refers to the degree of truncation of professional terms; the readability refers to the degree of acceptance of the layout of the second interface by the user; and the path consistency refers to whether the function of any operable component in the second interface is consistent with the corresponding operable component in the first interface.

8. The method of claim 7, wherein, generating the second interface information in the second language by using a genetic algorithm and an objective function according to the interface style information corresponding to the second language, the second layout information, and the second text information includes: generating a plurality of first candidate interface information according to the interface style information corresponding to the second language, the second layout information, and the second text information, and a genetic algorithm; selecting first optimal interface information from the plurality of first candidate interface information according to the objective function, and determining the first optimal interface information as the second interface information in the second language.

9. The method of claim 8, wherein, determining the first optimal interface information as the second interface information in the second language includes: adjusting the size, position, and interface style of any component in the first optimal interface information to generate a plurality of second candidate interface information under the condition that an iteration condition is met; the plurality of second candidate interface information includes the first optimal interface information; selecting optimal interface information from the plurality of second candidate interface information according to the objective function, and determining the optimal interface information as the second interface information.

10. The method of claim 7, wherein, The method further includes: adjusting the second interface information according to feedback information of the user; determining a first target result of the adjusted second interface information in the objective function; determining a second target result of the unadjusted second interface information in the objective function; if the difference between the first target result and the second target result is greater than a first threshold, determining that the feedback information of the user is invalid feedback.

Images