Method and system for generating personalized function page based on large model

Abstract

The application provides a large model-based personalized function page generation method and system, belonging to the technical field of large models. First, the function page requirement description input by the user through natural language is obtained, including function expectation and style preference. Then, the requirement analysis module of the large model is called for semantic analysis to obtain function requirement features and style preference features. Then, the function requirement features are used to search the preset component library to match the page component set, and the component display attributes are determined according to the style preference features. Then, the initial function page layout is generated according to the component set, the display attributes and the preset layout rules. Finally, the user interaction behavior data is monitored in real time, the page component display attributes and the layout structure are adjusted by inputting the large model dynamic adjustment module, and the personalized function page is obtained. Therefore, the function page meeting the personalized requirements of the user can be efficiently and accurately generated.

Description

Technical Field

[0001] This invention relates to the field of large model technology, and more specifically, to a method and system for generating personalized functional pages based on large models. Background Technology

[0002] In today's booming digital applications, the design of functional pages for various software and websites is crucial for improving user experience and meeting user needs. However, traditional methods of generating functional pages have many shortcomings.

[0003] On the one hand, traditional methods typically rely on professional developers to manually code and build functional pages. Developers need a deep understanding of user needs and translate that into concrete code logic—a process that is not only time-consuming and labor-intensive but also demands extremely high levels of professional skill. Furthermore, due to differences in the technical skills and comprehension abilities of different developers, the resulting pages vary significantly in terms of functional completeness and user experience.

[0004] On the other hand, while some existing automated page generation tools can simplify the development process to some extent, they often lack a precise grasp of users' personalized needs. These tools typically generate pages based on preset templates and rules, making it difficult to flexibly adjust them according to users' expectations for page functionality and presentation preferences. Users can only choose from a limited number of templates, failing to obtain truly personalized functional pages that meet their needs, thus affecting user experience and satisfaction. Furthermore, traditional methods lack effective monitoring and dynamic adjustment mechanisms for user interaction after page generation, making it impossible to optimize and improve the page based on actual user usage, resulting in poor usability and adaptability of the page. Summary of the Invention

[0005] In view of the aforementioned problems, and in conjunction with the first aspect of the present invention, embodiments of the present invention provide a method for generating personalized functional pages based on a large model, the method comprising:

[0006] Obtain the user's functional page requirement description input via natural language. The functional page requirement description includes the user's expectations for page functions and preferences for page presentation style.

[0007] The large model's requirement analysis module is invoked to perform semantic parsing on the functional page requirement description, thereby obtaining the user's functional requirement characteristics and style preference characteristics. The functional requirement characteristics reflect the user's expected page function type and interaction logic, while the style preference characteristics reflect the user's expected page color scheme, font style, and layout structure.

[0008] Based on the functional requirement features, a preset component library is retrieved and matched to obtain a set of page components corresponding to the functional requirements. At the same time, based on the style preference features, the display attributes of each component in the set of page components are determined. The display attributes include component size, position ratio and color parameters.

[0009] An initial functional page layout is generated based on the set of page components, the display attributes of each component, and the preset layout rules;

[0010] Real-time monitoring of user interaction data with the initial functional page; inputting the interaction data into the dynamic adjustment module of the large model; adjusting the display attributes of page components and page layout structure based on the interaction data to obtain a personalized functional page.

[0011] In another aspect, embodiments of the present invention also provide a system for generating personalized functional pages based on large models, including a processor and a machine-readable storage medium. The machine-readable storage medium is connected to the processor. The machine-readable storage medium is used to store programs, instructions, or code. The processor is used to execute the programs, instructions, or code in the machine-readable storage medium to implement the above-described method.

[0012] Based on the above, this embodiment of the invention obtains the user's natural language input of functional page requirement descriptions, enabling the user to express their needs more directly. The requirement parsing module of the large model performs semantic parsing of the requirement descriptions, accurately extracting the user's functional requirement features and style preference features. Based on the functional requirement features, a preset component library is retrieved and matched with the page component set. Simultaneously, the display attributes of the components are determined according to the style preference features, realizing intelligent selection and personalized customization of page components, improving the efficiency and accuracy of page generation. An initial functional page layout is generated based on the page component set, display attributes, and preset layout rules. Then, the user's interaction behavior data with the initial functional page is monitored in real time and input into the dynamic adjustment module of the large model. Based on the interaction behavior data, the display attributes of the page components and the page layout structure are dynamically adjusted, allowing the page to adaptively optimize according to the user's actual usage, continuously conforming to the user's needs and usage habits. This enables the efficient and accurate generation of pages that meet the user's personalized needs, significantly improving the user experience. Attached Figure Description

[0013] Figure 1 This is a schematic diagram of the execution flow of the personalized functional page generation method based on a large model provided in an embodiment of the present invention.

[0014] Figure 2 This is a schematic diagram of exemplary hardware and software components of the large-model-based personalized function page generation system provided in an embodiment of the present invention. Detailed Implementation

[0015] The present invention will now be described in detail with reference to the accompanying drawings. Figure 1 This is a flowchart illustrating a method for generating personalized functional pages based on a large model, according to an embodiment of the present invention. The following is a detailed description of this method.

[0016] Step S110: Obtain the user's functional page requirement description input via natural language. The functional page requirement description includes the user's expectations for page functions and preferences for page presentation style.

[0017] This embodiment uses an online learning platform as the application scenario. Users can input natural language descriptions through the interactive entry provided by the online learning platform. The natural language descriptions must fully cover two aspects: one is the expectation of the page functions, such as the expectation that the page has functions such as course resource management, student homework correction, and learning data statistics; the other is the preference for the page presentation style, such as the page's color scheme, font selection, and module layout.

[0018] Users with different roles will have different priorities. Teachers may be more concerned with the implementation of teaching management functions, students may be more concerned with access to learning resources and interactive communication functions, while parents may focus on the function of viewing students' learning progress.

[0019] Step S111: Receive the natural language requirement description submitted by the user through the input interface, which supports text input, speech-to-text input, and image with accompanying text description.

[0020] The input interface of the online learning platform consists of multiple functional modules. The text input module is represented by text input boxes on both the platform's web and mobile versions, where users can directly type text descriptions. The input boxes feature automatic line wrapping, word count suggestions, and a scroll bar that displays when the input reaches a certain length, allowing users to view the entire content.

[0021] The speech-to-text input module comprises a speech acquisition unit and a speech recognition unit. After the user clicks the speech input button, the speech acquisition unit activates microphone access and begins acquiring the user's speech information. During the acquisition process, a waveform graph is displayed in real time to provide feedback on the current recording status. The acquired speech information is transmitted to the speech recognition unit, which uses a deep learning-based speech recognition model to preprocess the speech signal, including noise reduction and framing operations, before converting the processed signal into corresponding text information.

[0022] The image-with-text description input module allows users to upload image files and enter supplementary descriptions of the image content in the text box below the image. The platform supports common image formats, and will verify the image format and size during the upload process. If the image does not meet the requirements, the user will be prompted to re-upload. After successful upload, the image and accompanying text description will be stored together.

[0023] Step S112: Identify the identity type of the input user. The identity types include students, teachers, and parents. Different identity types correspond to different focuses in the description of needs.

[0024] The platform identifies a user's identity type through their login information. When a user logs in, their account information is verified, and the platform retrieves the corresponding identity identifier from the user database. This identity identifier is set and stored during user registration and is uniquely associated with the account information.

[0025] When a user is not logged in, the input interface will prompt the user to log in first; otherwise, the user cannot submit the feature page description. Login status is monitored in real time. If the login status expires during input, the system will retain the current input and prompt the user to log in again.

[0026] Step S113: For student users, extract the functional expectations related to learning content from their input natural language requirement description; for teacher users, extract the functional expectations related to teaching management from their input natural language requirement description; for parent users, extract the functional expectations related to displaying student learning progress from their input natural language requirement description.

[0027] For student users, a combination of keyword matching and semantic analysis is used to extract relevant expressions. First, a set of keywords related to the learning content is set, such as "course videos," "exercises," and "knowledge point explanations." Then, the text input by the user is segmented into words, and the segmentation results are matched with the keyword set. At the same time, semantic analysis is used to determine the core meaning of the statement and filter out the expected expressions related to the learning content.

[0028] For teachers, a set of keywords related to teaching management is set up, such as "homework assignment", "grade entry", and "student attendance". Through similar word segmentation and matching processes, combined with semantic understanding, the expected expressions of functions related to teaching management are extracted.

[0029] For parents, a set of keywords related to displaying student learning progress is set, such as "learning progress", "exam scores", and "classroom performance". The corresponding expected descriptions are then extracted using the method described above.

[0030] Step S114: Redundant information filtering is performed on the extracted user requirement statements of different identity types, and the core requirement descriptions are retained as the functional page requirement descriptions input by users through natural language.

[0031] Redundant information filtering includes removing duplicate content and meaningless expressions. The system first performs text comparison on the extracted requirement expressions, identifies repeated sentences or paragraphs, and retains only one of them.

[0032] For meaningless expressions, such as interjections and repetitive exclamations, the system establishes a meaningless vocabulary database. It matches the words in the requirement description against the database, filtering out any successfully matched words. After this filtering process, the remaining content is determined as the core requirement description.

[0033] Step S120: Call the large model's requirement parsing module to perform semantic parsing processing on the functional page requirement description to obtain the user's functional requirement features and style preference features. The functional requirement features reflect the user's expected page function type and interaction logic, and the style preference features reflect the user's expected page color matching, font style, and layout structure.

[0034] After receiving the processed core requirement description, the large model's requirement parsing module initiates the semantic parsing process. This module contains multiple sub-modules, each responsible for parsing functional requirements and style preferences.

[0035] During the parsing process, the requirements parsing module performs deep semantic analysis on the text, identifying descriptions of page function types, interaction logic, color matching, font style, and layout structure, and then generating corresponding functional requirement features and style preference features.

[0036] Step S121: Input the functional page requirement description into the word vector conversion layer of the large model to generate a corresponding requirement word vector sequence. Each vector in the requirement word vector sequence corresponds to a word unit in the requirement description.

[0037] After receiving the functional page's requirement description text, the word vector conversion layer first preprocesses the text, including removing punctuation marks and unifying capitalization. Then, it segments the processed text into multiple lexical units, each corresponding to a word or phrase in the text.

[0038] For each lexical unit, the word vector transformation layer converts it into a fixed-dimensional vector through a pre-trained word embedding model. These vectors can capture the semantic information of the word, making words with similar meanings closer together in the vector space.

[0039] The generated sequence of required word vectors is arranged in the order of the word units in the original text, forming an ordered vector sequence.

[0040] Step S122: Call the context encoding module of the large model to perform semantic association modeling on the required word vector sequence, and generate an encoded vector containing contextual dependencies.

[0041] After receiving the required word vector sequence, the context encoding module models its semantic associations. This module adopts a Transformer architecture, using a multi-head attention mechanism and a feedforward neural network layer to capture contextual information.

[0042] Step S1221: Input the required word vector sequence into the multi-head attention layer of the context encoding module, and calculate the attention weight value between each word unit and other word units. The attention weight value is used to represent the semantic association strength between word units.

[0043] The multi-head attention layer maps the demand word vector sequence through multiple linear transformation matrices to obtain multiple different query matrices, key matrices, and value matrices. For each matrix combination, the dot product of the query matrix and the key matrix is ​​calculated, and then scaled to obtain the original attention weights.

[0044] The original attention weights are normalized using the softmax function to obtain the attention weight values ​​between each lexical unit and other lexical units. These weight values ​​reflect the degree of correlation between the various lexical units in the current context.

[0045] Step S1222: Based on the attention weight value, perform weighted aggregation processing on the required word vector sequence to generate a context enhancement vector for each word unit. The context enhancement vector integrates the semantic information of other word units whose relevance meets the preset relevance criteria.

[0046] Based on the attention weight values, the vector corresponding to each lexical unit in the required word vector sequence is weighted. The vector of each lexical unit is then summed with the vectors of other lexical units according to their attention weight values ​​to obtain the initial context enhancement vector.

[0047] The preset association criterion is a threshold for the attention weight value. For lexical units with weight values ​​below this threshold, their semantic information will not be included in the context enhancement vector of the current lexical unit. In this way, it is ensured that the context enhancement vector only incorporates the semantic information of lexical units with high relevance.

[0048] Step S1223: Input the context enhancement vector into the bidirectional recurrent layer for temporal association modeling to generate a temporal encoding vector containing semantic dependencies between context and background. Each dimension of the temporal encoding vector corresponds to the semantic features of a position in the required word vector sequence.

[0049] The bidirectional recurrent layer consists of forward recurrent units and backward recurrent units. The context enhancement vectors are input sequentially into the forward recurrent unit, which calculates the hidden state at the current time step based on the current input and the hidden state at the previous time step, capturing the positive temporal correlations between lexical units.

[0050] Meanwhile, the context enhancement vector is input into the backward recurrent unit in reverse order. The backward recurrent unit also calculates the hidden state at the current time step based on the current input and the hidden state at the previous time step, capturing the reverse temporal correlation between lexical units.

[0051] The hidden states of the forward and backward recurrent units at each position are concatenated to generate a temporal encoding vector, which contains the contextual semantic dependencies of the lexical units throughout the sequence.

[0052] Step S1224: Perform layer normalization on the time-series coding vector to obtain a standardized time-series coding vector.

[0053] Layer normalization is performed on each feature dimension of the temporal coding vector. The mean and variance of all vector elements in each feature dimension are calculated. Then, the mean is subtracted from each element and divided by the square root of the variance. This is then multiplied by the scaling parameter and the offset parameter is added to obtain the normalized temporal coding vector.

[0054] Layer normalization makes the distribution of temporal coding vectors more stable across different feature dimensions, which is beneficial for subsequent network layer processing.

[0055] Step S1225: The standardized temporal encoding vector is subjected to dimensionality transformation processing through the output layer of the context encoding module to generate an encoding vector with the same length as the required word vector sequence. Each element of the encoding vector contains the contextual semantic information of the corresponding lexical unit.

[0056] The output layer processes the standardized temporal encoding vector using a linear transformation. A linear transformation matrix is ​​used to convert the dimension of the standardized temporal encoding vector to match the number of lexical units in the desired word vector sequence, thus generating the encoding vector.

[0057] Each element in the encoding vector corresponds to a lexical unit in the required word vector sequence and contains the semantic information of that lexical unit in the entire context.

[0058] Step S123: Perform feature splitting on the encoding vector through the feature separation layer of the large model to separate the functional encoding sub-vectors related to functional requirements and the style encoding sub-vectors related to style preferences.

[0059] The feature separation layer contains two parallel convolutional neural network branches, used to extract the function encoding sub-vector and the pattern encoding sub-vector, respectively. The encoded vector is input to both branches simultaneously.

[0060] The branch used to extract the functional encoding sub-vector employs multiple convolutional kernels, which are trained to recognize semantic features related to functional requirements. Functional encoding sub-vectors are extracted from the encoding vector through convolution and pooling operations.

[0061] Another branch also uses multiple convolutional kernels to identify semantic features related to style preferences. After convolution and pooling operations, the style encoding sub-vectors are extracted.

[0062] Step S124: Perform functional dimension mapping processing on the functional encoding sub-vector to generate functional requirement features that include functional type identifiers, interaction logic descriptions, and priority ranking. The functional type identifiers are used to distinguish different page function categories, the interaction logic descriptions are used to explain the calling relationship between functions, and the priority ranking is used to indicate the importance of each function.

[0063] Functional dimension mapping is implemented through multiple fully connected layers. The functional encoding sub-vectors are processed sequentially through multiple fully connected layers. Each fully connected layer performs a non-linear transformation on the input through an activation function, gradually extracting higher-level functional features.

[0064] Step S1241: Call the function classifier to determine the category of the function encoding sub-vector and output the corresponding function type identifier. The function type identifier covers information display, operation interaction, data processing and feedback response.

[0065] The functional classifier consists of multiple fully connected layers and a softmax output layer. The functional encoding subvector is input into the fully connected layers, undergoes multiple nonlinear transformations, and then enters the softmax output layer.

[0066] The softmax output layer calculates the probability value for each function type and outputs the identifier corresponding to the function type with the highest probability value, i.e., the function type identifier. For example, if a function is determined to belong to the information display category, the corresponding information display category identifier is output.

[0067] Step S1242: Extract the semantic features related to the interaction from the function encoding sub-vector and generate an interaction logic description. The interaction logic description includes the function triggering condition, the function execution flow, and the function result return method. The function triggering condition is used to describe the user operation required to start the function. The function execution flow is used to describe the processing steps inside the function. The function result return method is used to describe the presentation form of the information after the function is completed.

[0068] By performing semantic analysis on the function encoding subvectors, interaction-related features are identified, including parts describing user operations, parts describing internal function processing steps, and parts describing results presentation.

[0069] Based on these characteristics, generate function trigger conditions, such as "the function will be launched after clicking the button"; function execution flow, such as "first verify user permissions, then load data, and finally process the data"; function result return method, such as "displayed in the middle of the page in list form", and combine these to form an interactive logic description.

[0070] Step S1243: Calculate the importance score of each function type based on the semantic intensity value in the function encoding sub-vector. The semantic intensity value is determined by the projection length of the function encoding sub-vector in the semantic activation direction. The larger the projection length, the higher the degree of emphasis the corresponding function is in the requirement description.

[0071] First, the semantic activation direction vector corresponding to each function type is determined. These vectors are obtained through training on a large amount of labeled data. The function encoding sub-vectors are then projected onto the semantic activation direction vectors of each function type, and the projection length is calculated. This projection length is the semantic intensity value.

[0072] Based on the semantic strength value, the importance score of each function type is calculated using a normalization method, so that the sum of the importance scores of all function types is 1. The higher the semantic strength value, the higher the corresponding importance score.

[0073] Step S1244: Sort each function type according to the importance score to generate a priority sort. The function type with a higher sequence position in the priority sort is the one that should be displayed first when the page is generated.

[0074] The functional types are ranked from highest to lowest importance to form a priority order. For example, if the information display category has the highest importance score, it will be ranked first in the priority order, followed by the operation and interaction category, and so on.

[0075] Step S1245: Combine the function type identifier, interaction logic description and priority sorting according to a preset format to generate the user's functional requirement features.

[0076] The default format is a structured data format, which maps the function type identifier, interaction logic description, and priority sorting to different fields in the structured data. For example, the function type identifier corresponds to the "Function Category" field, the interaction logic description corresponds to the "Interaction Rules" field, and the priority sorting corresponds to the "Display Priority" field. These elements are combined to form the functional requirement features.

[0077] Step S125: Perform style dimension mapping processing on the style encoding sub-vector to generate style preference features that include color system parameters, font attribute parameters and layout mode parameters. The color system parameters reflect the user's preference for the main color and auxiliary color of the page, the font attribute parameters reflect the user's preference for font type and size, and the layout mode parameters reflect the user's preference for the arrangement of page elements.

[0078] Style dimension mapping is also implemented through multiple fully connected layers. The style encoding sub-vectors undergo non-linear transformation by the fully connected layers to extract style features related to color, font, and layout.

[0079] For example, step S1251: parse the semantic features related to color in the style encoding sub-vector, determine the user's preferred main color range and auxiliary color matching rules, and generate color system parameters. The main color range is represented by coordinate intervals in the color space, and the auxiliary color matching rules are used to describe the brightness contrast and hue circle distance between the auxiliary color and the main color.

[0080] The color-related parts of the style encoding sub-vector are parsed to identify the main color information described by the user and convert it into coordinate ranges in the color space, such as the value ranges of the red, green and blue channels in the RGB color space, to form the main color range.

[0081] Simultaneously, the relationship between the auxiliary color and the primary color is analyzed to determine the range of brightness contrast and the range of hue distance on the color wheel, thus forming rules for auxiliary color matching. For example, the brightness contrast between the auxiliary color and the primary color is within a certain proportional range, and the hue distance on the color wheel is within a certain angular range.

[0082] Step S1252: Extract the semantic features related to the font from the style encoding sub-vector, determine the font type and font size range preferred by the user, and generate font attribute parameters. The font type includes serif fonts, sans-serif fonts and handwriting, and the font size range is represented by relative size levels.

[0083] Extract font-related semantic features from style-encoded subvectors, identify the font type mentioned by the user, such as sans-serif fonts, and determine the relative size range of the font size, such as medium-large to maximum size, to form font attribute parameters.

[0084] Step S1253: Analyze the semantic features related to layout in the style encoding sub-vector, determine the user's preferred page element arrangement and area division ratio, and generate layout mode parameters. The page element arrangement includes flow layout, grid layout and hierarchical layout. The area division ratio is used to describe the area proportion of different functional areas in the page.

[0085] Analyze the semantic features related to layout in the style-encoded sub-vectors to determine the user's desired arrangement of page elements, such as a grid layout. Simultaneously, parse the area proportion descriptions of different functional areas, such as the left functional area occupying two-thirds and the right functional area occupying one-third, to form layout pattern parameters.

[0086] Step S1254: The color system parameters, font attribute parameters, and layout mode parameters are associated and stored to generate the user's style preference features.

[0087] The color system parameters, font attribute parameters, and layout mode parameters are stored in key-value pairs. Each parameter is a value in a key-value pair, and the corresponding parameter name is the key. The combination forms the style preference feature.

[0088] Step S130: Based on the functional requirement features, retrieve the preset component library, match and obtain the set of page components corresponding to the functional requirements, and at the same time determine the display attributes of each component in the set of page components based on the style preference features. The display attributes include component size, position ratio and color parameters.

[0089] The pre-built component library stores a large number of pre-developed page components, each with corresponding functional tags and configurable attributes. The system searches the component library based on the functional type identifier in the functional requirements, matches the components that meet the requirements, and configures the display attributes of the components according to style preference characteristics.

[0090] Step S131: Extract the function type identifier and priority ranking from the function requirement features, and use the function type identifier as a search keyword to search the preset component library.

[0091] Extract the identifiers of each function type from the functional requirements characteristics, such as information display identifiers and operation interaction identifiers, and use these identifiers as search keywords to search in the index system of the preset component library.

[0092] Step S132: The preset component library stores multiple component units. Each component unit contains a function matching tag, basic structure parameters, and a set of configurable attributes. The function matching tag is used to match the function type identifier. The basic structure parameters are used to describe the core function implementation method of the component. The set of configurable attributes is used to represent the range of adjustable display attributes of the component.

[0093] The functional matching tags for component units are pre-defined and correspond to the functional type identifier. Basic structural parameters include the component's code structure, data interface specifications, and a list of dependent resources. The code structure clearly defines the component's implementation logic and module division. For example, the code structure of an information display component includes a data receiving module, a rendering module, and an update module. The data receiving module is responsible for obtaining display data from the backend, the rendering module presents the data to the page in a specific format, and the update module triggers a re-render when the data changes.

[0094] The data interface specification defines the format and protocol for data interaction between the component and the backend, including the type and format of request parameters and the structure of response data, ensuring that the component can correctly obtain and parse data. The dependency resource list lists the external resources required for the component to run, such as stylesheet files and script files, which are automatically imported when the component loads.

[0095] The configurable property set covers various adjustable parameters for the component's display. For color properties, it includes the settable range for primary color, secondary color, background color, etc.; font properties include optional values ​​for font type, font size, font weight, etc.; layout properties include the adjustment range for width, height, margins, positioning method, etc. Each configurable property has explicit value constraints to ensure that the component can still display and function correctly after adjustments.

[0096] Step S133: Calculate the semantic similarity between the function type identifier and the function matching tag of the component unit, and select the component units whose semantic similarity meets the preset conditions as candidate components.

[0097] First, the function type identifier and the function matching label of the component unit are converted into vector representations. A word embedding model is then used to process the function type identifier and function matching label, generating corresponding vectors. These vectors reside in the same vector space and reflect their semantic information.

[0098] Then, semantic similarity is determined by calculating the cosine similarity between the two vectors. The cosine similarity is calculated based on the angle between the vectors; the smaller the angle, the larger the cosine similarity value, indicating a closer semantic connection between the two vectors.

[0099] The calculated semantic similarity is compared with a preset condition, which is a semantic similarity threshold. If the semantic similarity between the function matching tag and the function type identifier of a component unit is greater than the threshold, the component unit is included in the candidate component list.

[0100] For example, in step S1331: the function type identifier is converted into the corresponding function vector, and the function matching label of the component unit is converted into the corresponding label vector. The function vector and the label vector are represented in the same vector space.

[0101] Both the function type identifier and the function matching label are transformed using the same pre-trained word embedding model. This word embedding model is trained on a large-scale text corpus and can map text information to a high-dimensional vector space, with semantically similar texts having a relatively close distance in the vector space.

[0102] For example, for the functional type identifier "information display", the word embedding model will convert it into a specific functional vector; for the functional matching label "component used to display course information" in the component unit, it will also be converted into a label vector, and the two vectors are in the same vector space, which facilitates subsequent similarity calculation.

[0103] Step S1332: Calculate the cosine similarity between the function vector and the label vector to obtain the semantic similarity value between the function type identifier and the function matching label.

[0104] The calculation process for cosine similarity is as follows: first, calculate the dot product of the function vector and the label vector, then calculate the magnitude of each vector, and finally divide the dot product by the product of the two magnitudes. The result is the cosine similarity value.

[0105] The value ranges from -1 to 1. The closer the value is to 1, the more similar the semantics of the function type identifier and the function matching tag are; the closer the value is to -1, the greater the semantic difference between the two.

[0106] Step S1333: Compare the semantic similarity value with the preset threshold. If the semantic similarity value meets the preset conditions, the corresponding component unit is included in the candidate component list.

[0107] The preset threshold is set according to the actual application scenario and requirements, and is generally between 0 and 1. When the calculated semantic similarity value is greater than or equal to the preset threshold, the functional matching tag and functional type identifier of the component unit are considered to be semantically similar enough to meet the preset conditions, and it is added to the candidate component list.

[0108] For example, if the preset threshold is 0.7, and the semantic similarity value of a component is 0.8, then that component will be included in the candidate component list; if the semantic similarity value is 0.6, then it will not be included.

[0109] Step S1334: Perform functional redundancy detection on the component units in the candidate component list. If two component units have completely overlapping core functions, retain the component unit with a semantic similarity value greater than the set similarity value, remove the component unit with a semantic similarity value not greater than the set similarity value, and use the component unit after redundancy detection as a candidate component.

[0110] Functional redundancy detection is achieved by analyzing the core functional descriptions in the basic structural parameters of the component unit. The core functional descriptions clearly define the primary functional purpose of the component unit, such as "displaying a course list" or "submitting assignments."

[0111] The core functional descriptions of all component units in the candidate component list are compared pairwise. If two component units have completely identical core functional descriptions, they are considered to have functional redundancy. In this case, the semantic similarity values ​​of these two component units with the functional type identifiers are compared, and the component unit with the higher semantic similarity value is retained, while the other is removed.

[0112] If two component units have the same semantic similarity value, one of them is randomly retained. After the above processing, there will be no component units in the candidate component list that are completely redundant in function.

[0113] Step S134: Convert the priority ranking into a numerical weight coefficient, which, together with semantic similarity, determines the component selection order. Perform a weighted ranking on the candidate components and select the top K candidate components in the ranking sequence to form a set of page components corresponding to the functional requirements.

[0114] Priority ranking reflects the importance of each function type. When converting it into numerical weight coefficients, the corresponding weight coefficients are assigned in descending order of priority. For example, the highest priority function type has a weight coefficient of 1.0, the second highest has 0.8, and so on. The specific values ​​are set according to the length and gradient of the ranking.

[0115] For each candidate component, its semantic similarity to the function type identifier is multiplied by the weight coefficient corresponding to that function type to obtain a weighted score. The candidate components are then sorted in descending order of their weighted scores, and the top K candidate components are selected to form the page component set. The value of K is determined based on the page's capacity and the complexity of the functional requirements, ensuring that the page does not appear cluttered due to an excessive number of components, while still meeting the core functional requirements.

[0116] Step S135: Extract the color system parameters, font attribute parameters, and layout mode parameters from the style preference features, and use them as the basis for configuring the component display attributes.

[0117] Color system parameters, font attribute parameters, and layout mode parameters are extracted from style preference features. Color system parameters include the primary color range and secondary color matching rules; font attribute parameters include font type and font size range; layout mode parameters include the arrangement of page elements and the proportion of area division.

[0118] These parameters will serve as the basis for subsequent adjustments to the component's display properties, ensuring that the component's display effect matches the user's style preferences.

[0119] Step S136: For each component in the page component set, adjust the component's color attributes according to the color system parameters, adjust the component's text display attributes according to the font attribute parameters, and adjust the component's size ratio and position parameters according to the layout mode parameters to generate the display attributes of each component. The display attributes include component size, position ratio, and color parameters.

[0120] For each component, adjust its properties one by one based on the extracted style parameters. During the adjustment process, refer to the component's own configurable property set to ensure that the adjusted properties are within the component's allowed configuration range.

[0121] For example, step S1361: For each component in the page component set, parse the color configuration items in its configurable attribute set, determine the main color value of the component according to the main color range in the color system parameters, determine the auxiliary color value of the component according to the auxiliary color matching rules, and assign the main color value and auxiliary color value to the corresponding color configuration items of the component to complete the adjustment of the component's color attributes.

[0122] Analyze the component's color configuration options to determine the specific locations of the main and secondary colors that need to be set, such as the title color, border color, and background color.

[0123] Based on the primary color range in the color system parameters, select a specific color value within that range as the main color value of the component. For example, select a specific blue value within the primary blue color range as the color of the title.

[0124] Based on the rules for matching auxiliary colors and the values ​​of the primary colors, determine the auxiliary color values. For example, if the rules for matching auxiliary colors require that the brightness and contrast of the auxiliary color and the primary color be within a certain range, then select auxiliary color values ​​that conform to that contrast range based on the brightness value of the primary color, and use them as the border color or background color, etc., and assign these color values ​​to the corresponding color configuration items.

[0125] Step S1362: Parse the font configuration items in the component's configurable attribute set, determine the component's text font based on the font type in the font attribute parameters, determine the component's text size based on the font size range, apply the text font and text size to the component's font configuration items, and complete the adjustment of the component's text display attributes.

[0126] The component's font configuration options are analyzed to clarify the configurable font and font size options for different text parts (such as title text, content text, and prompt text).

[0127] Based on the font type in the font attribute parameters, select the corresponding font for each text part of the component, such as a specific font in the sans-serif font category.

[0128] Based on the font size range, combined with the overall design of the component and the importance of the text, determine specific font sizes for different text parts. For example, select a larger font size within the font size range for title text and a medium font size for content text. Then apply the determined font and font size to the corresponding font configuration items.

[0129] Step S1363: Parse the layout configuration items in the component's configurable attribute set, determine the component's size ratio based on the area division ratio in the layout mode parameters, determine the component's position coordinates on the page based on the arrangement of page elements, set the size ratio and position coordinates to the component's layout configuration items, and complete the adjustment of the component's size ratio and position parameters.

[0130] Analyze the component's layout configuration options to understand the adjustable layout parameters such as the component's width, height, and positioning method.

[0131] The size proportion of a component within its functional area is determined based on the area division ratio in the layout mode parameters. For example, if a functional area occupies one-half of the page, and the component occupies one-third of the page within that functional area, then the component's overall size proportion on the page is one-sixth.

[0132] Based on the page element arrangement, such as a grid layout, determine the component's specific position coordinates within the grid, specifying its row and column. Set the determined size proportions and position coordinates into the component's layout configuration settings to complete the layout attribute adjustment.

[0133] Step S1364: Integrate the adjusted color attributes, text display attributes, size ratio, and position parameters to generate display attributes that include component size, position ratio, and color parameters.

[0134] The adjusted color attributes, text display attributes, size ratios, and position parameters are summarized and organized, and combined into the component's display attributes according to a preset format. The component size is determined by the size ratio, the position ratio reflects the component's space share on the page, and the color parameters include information such as the main color value and auxiliary color values.

[0135] Each component's display attributes are associated with that component, facilitating the generation of subsequent page layouts.

[0136] Step S137: Associate and bind the page component collection with the display attributes of each component so that the display attributes of each component are adapted to their priority order in the functional requirements features.

[0137] By establishing mapping relationships, each component in the page component collection is associated and bound to its corresponding display attributes. During the binding process, the display attributes are fine-tuned based on the priority order of the component's corresponding functional type.

[0138] For example, components corresponding to higher-priority functional types may have their size percentage in the display attributes increased appropriately, and their positions may be closer to the visual focus area of ​​the page to highlight their importance and ensure that the display effect of the component matches the priority of the function.

[0139] Step S140: Generate an initial functional page layout based on the set of page components, the display attributes of each component, and the preset layout rules.

[0140] The preset layout rules include principles for dividing functional areas, component placement guidelines, and interaction logic adaptation. Based on these rules, and combined with the collection of page components and the display attributes of each component, the system plans and designs the page layout, determining the final position, size, and style of each component on the page, thus forming the initial functional page layout.

[0141] Step S141: Extract the function type identifier and priority order of each component in the page component set. Based on the function area division principle in the preset layout rules, divide the page into multiple function areas, with each function area corresponding to one or more function type identifiers.

[0142] The functional area division principle in the preset layout rules divides the page into different functional areas based on the relevance of functional types and the visual layout habits of the page. For example, information display functions are concentrated in the left area of ​​the page, and operation and interaction functions are concentrated in the right area, or data processing functions are placed in the upper area of ​​the page, and feedback and response functions are placed in the lower area.

[0143] Based on the functional type identifiers of each component, the functional type corresponding to each functional area is determined, ensuring that components belonging to the same or related functional types can be assigned to the same functional area. Simultaneously, based on priority ranking, functional areas with larger areas or more advantageous locations are assigned to functional types with higher priority.

[0144] Step S142: Based on the priority of each component, assign the K components with the highest priority in the priority sorting sequence to functional areas on the page with visual weights that meet the first preset level, and assign the M components with the lowest priority in the priority sorting sequence to functional areas on the page with visual weights that meet the second preset level. The visual weights are determined by the position and size of the functional areas on the page, and the visual weights of the first preset level are greater than the visual weights of the second preset level.

[0145] The first preset level of visual weight corresponds to a prominent and large area on the page, such as the upper central area of ​​the page; the second preset level corresponds to a relatively minor and smaller area, such as the edge or bottom area of ​​the page.

[0146] Based on component priority, the top K components are assigned to functional areas with a visual weight of the first preset level. These components are typically the most frequently used or most important functional components. The bottom M components are assigned to functional areas with a visual weight of the second preset level. The values ​​of K and M are determined based on the size of the functional areas and the number of components to ensure that each functional area can reasonably accommodate the assigned components.

[0147] Step S143: Based on the size ratio and position parameters in the display attributes of each component, determine the specific placement position and display size of each component in the corresponding functional area so that there is no overlap between components.

[0148] Within each functional area, the actual display size of each component is calculated based on the size percentage in the component's display attributes. Combined with position parameters, the specific coordinates of the component within the functional area are determined.

[0149] When determining the location, a collision detection mechanism is used to check whether newly placed components overlap with existing components. If overlap exists, the position parameters of the components are adjusted, such as fine-tuning the coordinate values, until all components are clearly displayed within the functional area and there is no overlap between them.

[0150] Step S144: Analyze the functional relationships between components. Based on the principle of interaction proximity in the preset layout rules, place components with direct interaction relationships in adjacent positions to shorten the user operation path.

[0151] The functional relationships between components are reflected in their interaction logic descriptions. For example, the output of one component is the input of another component, or the operation of one component will trigger the response of another component.

[0152] The principle of proximity in the default layout rules requires that components with direct interactive relationships should be placed in adjacent positions as much as possible to reduce the distance users need to jump between pages and move during operation.

[0153] For example, step S1441: obtain the interaction logic description of each component in the page component collection, extract the function call relationship between components, and generate a component interaction relationship diagram. In the component interaction relationship diagram, each node represents a component, and the connection between nodes represents the direct interaction relationship between components.

[0154] From the interaction logic description of each component, filter out the parts that involve other components, such as "component A will be called after this component completes its operation" or "the output data of component B will be passed to this component", in order to determine the functional call relationship between components.

[0155] Treating each component as a node, draw lines between two nodes that have a direct interaction relationship to form a component interaction diagram. The lines can indicate the direction and type of the interaction, such as one-way calls or two-way data transmission.

[0156] Step S1442: Calculate the interaction frequency between each component based on the component interaction relationship diagram. The higher the interaction frequency value, the closer the relationship between the components.

[0157] Interaction frequency refers to the number of direct interactions between two components within a preset time range or based on historical interaction data. Interaction frequency is obtained by analyzing component interaction logs or simulating interaction scenarios to count the number of interactions between each pair of components.

[0158] For example, if component A and component B interact directly 10 times in a day, and component C and component D interact 5 times, then the interaction frequency of component A and component B is greater than the interaction frequency of component C and component D, indicating that A and B are more closely related.

[0159] Step S1443: According to the interaction proximity principle in the preset layout rules, mark the component pairs whose interaction frequency meets the preset threshold for adjacent placement.

[0160] The preset threshold is set based on the compactness of the page layout and the requirements of interaction efficiency. When the interaction frequency between two components is greater than or equal to the preset threshold, it is considered that the two components need to be placed in adjacent positions according to the principle of interaction proximity, and they are marked as adjacent.

[0161] For example, if the preset threshold is 8 times and the interaction frequency of component A and component B is 10 times, then they will be marked as adjacent.

[0162] Step S1444: When assigning component positions within a functional area, prioritize component pairs with adjacent placement marks and arrange them within the same or adjacent functional sub-areas so that the physical distance between components is positively correlated with the frequency of interaction.

[0163] The functional area is further divided into functional sub-areas, each with a small area to facilitate the centralized placement of closely related components. When allocating component locations, component pairs marked with adjacent placement are processed first and arranged within the same functional sub-area. If space is insufficient, they are arranged within adjacent functional sub-areas.

[0164] Components with higher interaction frequency are placed closer together in terms of physical distance, reflecting the positive correlation between physical distance and interaction frequency and improving the convenience of user operation.

[0165] Step S1445: Calculate the operation path length between adjacent component pairs. If the operation path length does not meet the preset standard, adjust the position parameters of the components until the operation path length meets the preset standard.

[0166] Operation path length refers to the distance a user needs to move from one component to another to perform an operation, typically measured in pixels on the screen. The default standard is a maximum allowed operation path length value.

[0167] For adjacent component pairs, calculate the operation path length between them. If the length is greater than the preset standard, fine-tune the position parameters of one or two components, such as changing the coordinate values, reducing the distance between them, and recalculate the operation path length until it is less than or equal to the preset standard.

[0168] Step S1446: Record the final placement of components and their corresponding interaction relationships as the interaction association attributes of the initial functional page layout.

[0169] Record the final coordinates of each component on the page, its functional area and sub-areas, as well as the interaction relationships and adjacent placement of components, to form the interaction association attributes of the initial functional page layout. These interaction association attributes can reflect the spatial relationships and functional interaction relationships between components.

[0170] Step S145: Perform global coordination processing on the color parameters in the display attributes of each component so that the color matching of different components conforms to the color system parameters in the style preference characteristics, forming a unified page color style.

[0171] Although the color attributes of each component have been adjusted according to the color system parameters, there may be color inconsistencies between different components, thus requiring global coordination. First, the color parameters of all components, including primary and secondary color values, are extracted and converted to a unified color space for comparison and adjustment.

[0172] Next, based on the color system parameters in the style preference features, especially the primary color range and secondary color matching rules, the color parameters of each component are verified one by one. For components whose main color value exceeds the primary color range, their main color value is adjusted to the primary color range, while ensuring that the visual difference between the adjusted color and the original color is within an acceptable range.

[0173] For secondary colors, check whether their brightness contrast and hue wheel distance from the corresponding main component color conform to the secondary color matching rules. If not, adjust the secondary color value according to the rules. For example, when the brightness contrast is lower than the rule requirement, increase the brightness difference between the secondary color and the main color; when the hue wheel distance does not conform, adjust the hue of the secondary color so that its hue wheel distance from the main color is within the specified range.

[0174] During the adjustment process, the color scheme between adjacent components also needs to be considered. For components that are functionally closely related and placed next to each other, ensure that there is sufficient differentiation between their colors while avoiding jarring color clashes. For example, if two adjacent components both use a primary color as their main color, the differentiation can be enhanced by adjusting their secondary colors or brightness.

[0175] After adjusting the color parameters of all components, review the overall color effect of the page to ensure a unified color style that conforms to the color system parameters in the style preference characteristics. If there are any areas with inconsistent colors, make further fine-tuning until a unified and harmonious page color style is achieved.

[0176] Step S146: Integrate the placement, display size, color parameters, and functional area division of all components to generate the initial functional page layout.

[0177] Summarize the information obtained from the above steps, including component placement, display size, color parameters, and functional area division. Based on the page layout hierarchy, first determine the position and extent of each functional area on the page, and then arrange the components within each functional area according to their final placement and display size.

[0178] Simultaneously, the color parameters of each component are applied to the corresponding component to ensure that the color display of the components conforms to the globally coordinated result. This information is integrated into a complete page layout data structure, which includes the overall size of the page, details of the division of functional areas, specific attributes of each component (position coordinates, size, color parameters, etc.), and interactive relationship attributes between components.

[0179] This integration process generates an initial functional page layout that meets the user's functional needs and style preferences, providing the user with a preliminary functional page display.

[0180] Step S150: Monitor the user's interaction behavior data with the initial functional page in real time, input the interaction behavior data into the dynamic adjustment module of the large model, and adjust the display attributes of the page components and the page layout structure based on the interaction behavior data to obtain a personalized functional page.

[0181] After the initial functional page is displayed to the user, the system activates a real-time monitoring mechanism to capture various user interactions on the page. Once the interaction data is collected, it is sent to the dynamic adjustment module of the large model for analysis and processing. Based on the analysis results, the display attributes and layout structure of the page components are adjusted to make the page more in line with the user's actual usage habits and needs, ultimately forming a personalized functional page.

[0182] Step S151: Deploy the behavior monitoring component to collect real-time interaction behavior data between the user and the initial function page. The interaction behavior data includes the number of component clicks, the duration of component dwell, the page scrolling trajectory, and the operation sequence.

[0183] An action monitoring component is embedded in the code of the initial functional page. This component can listen for various user action events on the page. When a user clicks on a component, the monitoring component records the component's identifier and the time of the click, accumulating the number of component clicks. When the user's mouse or touch focus hovers over a component, the monitoring component records the start and end times of the hover, calculating the duration of the hover.

[0184] For page scrolling, the monitoring component listens to page scrolling events and records information about changes in page position during scrolling, including scrolling direction, speed, and the page area reached, forming page scrolling trajectory data. The operation sequence records a series of user actions performed on the page in chronological order, such as clicking a component, scrolling the page, and then clicking another component. Each operation includes information such as the operation type, the components involved, and the operation time.

[0185] The collected interaction data is transmitted in real time to the data storage module for temporary storage, and undergoes preliminary format conversion to meet the data format requirements for subsequent processing. During the data collection process, interaction data involving user privacy is processed using data anonymization techniques, such as obfuscating the user's operation time and removing sensitive information that may identify the user, ensuring that data collection complies with privacy protection requirements.

[0186] Step S152: Preprocess the interactive behavior data by converting the number of component clicks into a click frequency parameter, the component dwell time into a dwell percentage parameter, the page scroll trajectory into a regional attention parameter, and the operation sequence into an interaction path parameter.

[0187] For component click counts, the number of clicks within a preset time interval is counted, and then divided by the time interval to obtain the number of clicks per unit time, i.e., the click frequency parameter. This click frequency parameter reflects the user's click activity on the component.

[0188] The preprocessing of component dwell time involves dividing the dwell time of each component by the total dwell time of the user on the entire page to obtain the dwell percentage parameter. This dwell percentage parameter represents the proportion of time the user spends on that component, which can reflect the user's attention to the component.

[0189] When processing page scrolling, the page is divided into multiple equal areas. The time users spend in each area and the number of times they scroll through that area are counted. Based on this data, the attention score of each area is calculated to form the area attention parameter. The higher the score, the more attention the area receives from users.

[0190] For the operation sequence, the components and page locations involved in each operation are extracted in chronological order. This information is then converted into nodes and paths. Nodes represent components or page areas, and paths represent the order and jump relationships of operations, thus forming interactive path parameters. These interactive path parameters can reflect the user's operation flow and path on the page.

[0191] In the preprocessed parameters, each dimension corresponds to a different component or page area, and the values ​​in the vector reflect the corresponding parameter values.

[0192] Step S153: Input the preprocessed click frequency parameters, dwell time percentage parameters, regional attention parameters, and interaction path parameters into the dynamic adjustment module of the large model to generate user behavior feature vectors.

[0193] The dynamic adjustment module of the large model receives the preprocessed parameters and inputs them into the input layer of the dynamic adjustment module. The input layer concatenates these vectors to form a comprehensive input vector that contains multifaceted information about user interaction behavior.

[0194] The input vector is processed through multiple hidden layers in a dynamic adjustment module. These hidden layers employ a fully connected neural network structure, using activation functions to perform non-linear transformations on the input, progressively extracting deep features of user behavior. The output of each hidden layer serves as the input to the next hidden layer. After multiple feature extractions and transformations, the output layer ultimately generates the user behavior feature vector.

[0195] The user behavior feature vector is a high-dimensional vector, with each dimension corresponding to a specific feature of user behavior, which can comprehensively reflect the user's behavioral habits and preferences when using the initial functional page.

[0196] Step S154: Analyze the user behavior feature vector through the demand change recognition layer of the dynamic adjustment module to determine whether there is a deviation between the user's current actual demand and the initial functional demand features. If there is a deviation, generate a demand adjustment coefficient.

[0197] The requirement change identification layer includes a feature comparison unit and a coefficient generation unit. The feature comparison unit receives user behavior feature vectors and initial functional requirement features, and transforms them into the same feature space for comparison and analysis.

[0198] For example, step S1541: extract the priority ranking from the initial functional requirement features as the baseline priority sequence, and convert the click frequency parameter and dwell percentage parameter in the user behavior feature vector into the actual attention sequence.

[0199] Priority ranking is extracted from the initial functional requirements characteristics. This priority ranking reflects the importance of each functional type initially expected by the user. It is then converted into a numerical baseline priority sequence, where each element corresponds to an initial importance score for a functional type.

[0200] For the user behavior feature vector, the click frequency parameter and dwell time percentage parameter are extracted. These two parameters are then weighted and fused. The weights are set according to the degree of influence of the two parameters on user attention. After fusion, the actual attention score corresponding to each function type is obtained, forming an actual attention sequence.

[0201] Step S1542: Calculate the deviation between the benchmark priority sequence and the actual attention sequence, wherein the deviation is represented by the sum of squared differences between the corresponding elements in the sequence.

[0202] The difference between corresponding elements in the baseline priority sequence and the actual attention sequence is calculated by subtracting them pairwise. Then, each difference is squared, and all squared values ​​are summed to obtain the deviation value. The magnitude of this deviation value reflects the degree of difference between the user's current actual needs and the initial functional requirements.

[0203] Step S1543: Compare the deviation value with a preset deviation threshold. If the deviation value is greater than the preset deviation threshold, it is determined that there is a deviation between the user's current actual needs and the initial functional requirements.

[0204] The preset deviation threshold is a critical value determined based on a large amount of user behavior data and page adjustment experience. It is used to judge whether the deviation is significant. When the calculated deviation value is greater than the threshold, it indicates that the user's actual behavior reflects a significant difference from the initial needs, and the page needs to be adjusted. If the deviation value is less than or equal to the preset deviation threshold, it is considered that the user's current needs are basically consistent with the initial needs, and no adjustment is required.

[0205] Step S1544: Determine the value of the demand adjustment coefficient based on the magnitude of the deviation value. The larger the deviation value, the larger the value of the demand adjustment coefficient. The demand adjustment coefficient is used to represent the magnitude of the adjustment.

[0206] A mapping relationship is established between the demand adjustment coefficient and the deviation value. This mapping relationship ensures that the larger the deviation value, the larger the demand adjustment coefficient. For example, when the deviation value is within a certain range, the demand adjustment coefficient takes a value within the corresponding range and increases monotonically as the deviation value increases. This method ensures that the larger the deviation, the greater the page adjustment, thus better adapting to the user's actual needs.

[0207] Step S1545: Based on the difference between the actual attention sequence and the baseline priority sequence, determine the adjustment direction of the demand adjustment coefficient. If the element value in the actual attention sequence is greater than the corresponding element value in the baseline priority sequence, the adjustment direction is to enhance the attributes of the corresponding component. If the element value in the actual attention sequence is less than the corresponding element value in the baseline priority sequence, the adjustment direction is to weaken the attributes of the corresponding component.

[0208] Compare the element values ​​at corresponding positions in the actual attention sequence and the baseline priority sequence one by one. When the element value in the actual attention sequence is greater than the corresponding element value in the baseline priority sequence, it indicates that the user's actual attention to the function type corresponding to that element is higher than the initial expectation. Therefore, the adjustment direction is to enhance the attributes of the component corresponding to that function type. Conversely, when the element value in the actual attention sequence is less than the corresponding element value in the baseline priority sequence, the adjustment direction is to weaken the attributes of the corresponding component.

[0209] Step S1546: Integrate the values ​​and adjustment directions of the requirement adjustment coefficients to generate requirement adjustment coefficients to guide the adjustment of component display attributes.

[0210] The value of the demand adjustment coefficient and the adjustment direction are combined into a structured parameter. This parameter contains both the adjustment magnitude and the adjustment direction information, which can clearly guide the subsequent adjustment operations of the component display attributes.

[0211] Step S155: Adjust the display attributes of page components based on the required adjustment coefficient, increase the size ratio of components whose click frequency parameter meets the first preset condition, enhance the color contrast of components whose dwell time parameter meets the second preset condition, and optimize the component layout of areas whose regional attention parameter meets the third preset condition.

[0212] Based on the adjustment range and direction indicated by the demand adjustment coefficient, the display attributes of page components are adjusted in a targeted manner to improve the matching degree between the page and the actual needs of users.

[0213] For example, in step S1551: For components whose click frequency parameters meet the first preset conditions, the size ratio of the component is increased according to the value and direction of the adjustment coefficient. The increase is determined by the relative value of the click frequency and the adjustment coefficient, and does not exceed the maximum allowable size ratio of the component.

[0214] The first preset condition is that the click frequency parameter is higher than a preset click frequency threshold, which is determined based on the average click frequency of page components. For components that meet this condition, their relative click frequency is calculated, which is the ratio of the component's click frequency to the average click frequency of all components.

[0215] Multiply the relative click frequency by the required adjustment coefficient to obtain the increase in size percentage. Based on the adjustment direction (enhancement), add the increase to the component's current size percentage to obtain the adjusted size percentage. Simultaneously, ensure that the adjusted size percentage does not exceed the maximum allowed size percentage for the component in the page layout; if it does, use the maximum allowed percentage as the adjusted size percentage.

[0216] Step S1552: For components whose dwell percentage parameter meets the second preset condition, adjust the value and direction of the adjustment coefficient according to the requirements to improve the color contrast of the component. The adjustment of color contrast is achieved by increasing the brightness difference between the main color and the background color. The increase in color contrast is determined by the dwell percentage parameter and the required adjustment coefficient, and it must comply with the color matching rules.

[0217] The second preset condition is that the dwell percentage parameter is higher than a preset dwell percentage threshold, which is determined based on the average dwell percentage of all components. For components that meet this condition, their dwell percentage parameter is multiplied by the value of the required adjustment coefficient to obtain the improvement in color contrast.

[0218] Color contrast is improved by increasing the difference in brightness between the component's primary color and the background color. Specifically, depending on the desired increase, the brightness of the background color is appropriately reduced or the brightness of the primary color is increased, thus widening the difference in brightness between the two. Throughout the adjustment process, the color matching rules in the style preference features are consistently followed to ensure that the adjusted color contrast meets the rule requirements and avoids color clashes.

[0219] Step S1553: Analyze the regional attention parameters, determine the regions on the page whose user attention meets the third preset condition, and adjust the layout of the components in the region to reduce the blank areas between components while maintaining the interactive relationship between components.

[0220] The third preset condition is that the regional attention parameter is higher than a preset attention threshold, which is determined based on the average attention of each region on the page. For regions that meet this condition, the current layout of the components within that region is analyzed, and the size of the blank areas between the components is measured.

[0221] While maintaining the interactive relationships and preventing overlap between components, the positions of the components can be fine-tuned to reduce the blank areas between them, making the component layout more compact. For example, the spacing between adjacent components can be appropriately reduced, or the arrangement of components can be adjusted to make full use of the available space.

[0222] Step S1554: For components whose element values ​​in the actual attention sequence are less than the corresponding element values ​​in the baseline priority sequence, adjust the value and direction of the coefficient according to the requirements, reduce the size ratio of the component or reduce its color contrast, and reduce its visual weight on the page.

[0223] For these components, the reduction in size ratio or color contrast is determined by adjusting the coefficient based on requirements. When reducing size ratio, the corresponding reduction amount is subtracted from the current size ratio to ensure that the adjusted size ratio is not lower than the minimum size ratio required for the component to display normally. When reducing color contrast, the brightness difference between the main color and the background color is reduced to make the component less prominent on the page, thereby reducing its visual weight.

[0224] Step S1555: After the adjustment is completed, the display attributes of each component are checked for consistency so that the adjusted size ratio and color parameters conform to the overall style of the style preference characteristics and avoid visual conflicts.

[0225] Check whether the adjusted component sizes are within the limits allowed by the page layout, and whether the size ratios between components are harmonious. For color parameters, check whether they conform to the color system parameters in the style preference features, whether the color matching between components is harmonious, and whether there are any glaring or discordant issues.

[0226] If any non-compliance is found, the display attributes of the corresponding components will be fine-tuned again until all components meet the requirements, ensuring a consistent overall page style and no visual conflicts.

[0227] Step S156: Analyze the smoothness of user operation based on the interaction path parameters. If there is an operation path length that meets the fourth preset condition, adjust the position parameters of the relevant components, optimize the page layout structure, and shorten the interaction path.

[0228] By analyzing the interaction path parameters, the smoothness of user operations is evaluated. For cases where the operation path is too long, the component positions are adjusted to optimize the page layout and improve user operation efficiency.

[0229] Step S1561: Parse the interaction path parameters, extract the sequence of steps the user takes to complete a specific operation task on the page and the jump trajectory between each step, and generate an operation path diagram.

[0230] Identify specific user tasks from the interaction path parameters, such as "submit assignment and view grades", and extract the corresponding step sequence. Each step corresponds to an operation, such as "click the assignment submission component", "fill in the assignment information", "click the submit button", "click the grade query component", etc.

[0231] Based on the location of the components involved in each step, draw the jump trajectory between each step to form an operation path diagram. In the operation path diagram, nodes represent steps, and lines represent jump trajectories. The length of the lines reflects the jump distance.

[0232] Step S1562: Calculate the jump distance between each step in the operation path diagram. The jump distance is represented by the pixel distance between the center points of the components. Convert the jump distance into a proportional distance relative to the overall page size. Add up the proportional distances between all steps to obtain the total length of the operation path.

[0233] For each jump trajectory in the operation path diagram, determine the coordinates of the center points of the components involved in the steps at both ends of the trajectory, and calculate the straight-line pixel distance between the two points, which is the jump distance.

[0234] Using the overall width and height of the page as a reference, the jump distance is converted into a proportional distance, i.e., the jump distance is divided by the overall page size (width or height, depending on the jump direction). The proportional distances between all steps are added together to obtain the total operation path length, which reflects the length of the operation path required for the user to complete a specific task.

[0235] Step S1563: Compare the total length of the operation path with a preset path threshold. If the total length of the operation path is greater than the preset path threshold, it is determined that there is an operation path that is too long.

[0236] The preset path threshold is determined based on the average path length for users to complete similar tasks. When the total path length exceeds this threshold, it indicates that the path for the user to complete the task is too long, the operation is not smooth enough, and the position of relevant components needs to be adjusted.

[0237] Step S1564: Identify the consecutive steps with the longest jump distance in the operation path, and determine the relevant components involved in the consecutive steps.

[0238] In the operation path diagram, identify the two or more consecutive steps with the longest jump distances; these steps constitute the bottleneck sections of the operation path. Determine the components involved in these steps; these are the components whose positions need to be adjusted.

[0239] Step S1565: Based on the interaction proximity principle in the preset layout rules, adjust the position parameters of the relevant components, and place the components involved in the consecutive steps with the longest jump distance close to each other to reduce the jump distance between them.

[0240] Based on the principle of proximity in interaction, the positions of related components are adjusted to bring them as close as possible. For example, if the jump distance between two components is the longest, they can be moved to adjacent positions on the page or to the same functional sub-area to shorten the jump distance between them.

[0241] During the adjustment process, it is necessary to ensure that moving a component does not affect the normal layout and functionality of other components, while maintaining the interaction and relationship between the component and other related components.

[0242] Step S1566: Recalculate the total length of the adjusted operation path. If the total length of the operation path is still greater than the preset path threshold, then repeatedly adjust the position parameters of the relevant components until the total length of the operation path is less than or equal to the preset path threshold.

[0243] After adjusting the positions of the relevant components, recalculate the total length of the operation path according to the method in step S1562 and compare it with the preset path threshold. If it still does not meet the requirements, adjust the position parameters of the relevant components again. If necessary, expand the adjustment range to involve more relevant components to ensure that the total length of the operation path meets the requirements of the preset path threshold. During repeated adjustments, it is necessary to maintain the interactive relationship between components and take into account the rationality and aesthetics of the overall page layout to avoid page structure chaos due to excessive adjustments.

[0244] Step S1567: Record the adjusted component position parameters and update the page layout structure to optimize the user operation path.

[0245] The component placement parameters, after multiple adjustments, were meticulously recorded, including the specific coordinates and size proportions of each component. Based on these records, the page layout structure was updated, and the placement and display size of each component on the page were redefined. The updated page layout structure effectively shortens the user's operation path, improves the smoothness and efficiency of user operations, and makes personalized function pages more in line with the user's actual usage habits.

[0246] Following the processing in step S150 above, the initial functional page is dynamically adjusted based on user interaction data, ultimately resulting in a functional page that meets the user's personalized needs. This functional page not only satisfies the user's initial functional requirements and style preferences but is also optimized based on the user's actual operational behavior, thus improving the user experience.

[0247] The entire method for generating personalized feature pages based on a large model also involves protecting user privacy and sensitive data. When collecting user interaction data and other privacy-sensitive data, data anonymization techniques are used to anonymize user personal identification information and remove information that can be directly linked to the user's identity. Simultaneously, encrypted transmission technology is employed to ensure that data is not leaked during transmission. In the data storage stage, access control mechanisms are implemented, allowing only authorized personnel to access relevant data, and data access behavior is logged for security auditing. These technical measures effectively protect user privacy and prevent the leakage of sensitive privacy data.

[0248] Figure 2 The illustration shows exemplary hardware and software components of a large-model-based personalized functional page generation system 100 that can implement the ideas of this application, according to some embodiments of this application. For example, processor 120 can be used in the large-model-based personalized functional page generation system 100 and to perform the functions in this application.

[0249] For example, a large-model-based personalized page generation system 100 may include a network port 110 connected to a network, one or more processors 120 for executing program instructions, a communication bus 130, and various forms of storage media 140, such as a disk, ROM, or RAM, or any combination thereof. Exemplarily, the large-model-based personalized page generation system 100 may also include program instructions stored in ROM, RAM, or other types of non-transitory storage media, or any combination thereof. The methods of this application can be implemented according to these program instructions. The large-model-based personalized page generation system 100 also includes an I / O interface 150 between the computer and other input / output devices.

[0250] Furthermore, this embodiment of the invention also provides a readable storage medium, wherein computer-executable instructions are preset in the readable storage medium, and when the processor executes the computer-executable instructions, the above-described method for generating personalized functional pages based on a large model is implemented.

[0251] It should be noted that, in order to simplify the description of the present invention and thus help to understand one or more embodiments of the invention, multiple features may sometimes be grouped into one embodiment, drawing or description thereof in the foregoing description of the embodiments of the present invention.

Claims

1. A method for generating personalized functional pages based on a large model, characterized in that, The method includes: Obtain the user's functional page requirement description input via natural language. The functional page requirement description includes the user's expectations for page functions and preferences for page presentation style. The large model's requirement analysis module is invoked to perform semantic parsing on the functional page requirement description, thereby obtaining the user's functional requirement characteristics and style preference characteristics. The functional requirement characteristics reflect the user's expected page function type and interaction logic, while the style preference characteristics reflect the user's expected page color scheme, font style, and layout structure. Based on the functional requirement features, a preset component library is retrieved and matched to obtain a set of page components corresponding to the functional requirements. At the same time, based on the style preference features, the display attributes of each component in the set of page components are determined. The display attributes include component size, position ratio and color parameters. An initial functional page layout is generated based on the set of page components, the display attributes of each component, and the preset layout rules; Real-time monitoring of user interaction data with the initial functional page; inputting the interaction data into the dynamic adjustment module of the large model; adjusting the display attributes of page components and page layout structure based on the interaction data to obtain a personalized functional page. The step of retrieving a preset component library based on the functional requirement features, matching it to obtain a set of page components corresponding to the functional requirements, and simultaneously determining the display attributes of each component in the set of page components based on the style preference features includes: Extract the function type identifier and priority ranking from the functional requirement features, and use the function type identifier as a search keyword to search the preset component library; the preset component library stores multiple component units, each component unit contains a function matching tag, basic structure parameters and a configurable attribute set. The function matching tag is used to match with the function type identifier, the basic structure parameters are used to describe the core function implementation method of the component, and the configurable attribute set is used to represent the range of adjustable display attributes of the component. Calculate the semantic similarity between the function type identifier and the function matching tag of the component unit, and filter out the component units whose semantic similarity meets the preset conditions as candidate components; The priority ranking is converted into a numerical weight coefficient, which, together with semantic similarity, determines the component selection order. The candidate components are then weighted and sorted, and the top K candidate components in the sorted sequence are selected to form a set of page components that correspond to the functional requirements. Extract the color system parameters, font attribute parameters, and layout mode parameters from the style preference features, and use them as the basis for configuring the component display attributes; For each component in the page component set, adjust the component's color attributes according to the color system parameters, adjust the component's text display attributes according to the font attribute parameters, and adjust the component's size ratio and position parameters according to the layout mode parameters to generate the display attributes of each component. The display attributes include component size, position ratio, and color parameters. Associating and binding the page component collection with the display attributes of each component ensures that the display attributes of each component are matched with their priority order in the functional requirements.

2. The method for generating personalized functional pages based on a large model according to claim 1, characterized in that, The description of the functional page requirements for obtaining user input via natural language includes: Receive natural language request descriptions submitted by users through an input interface, which supports text input, speech-to-text input, and image-with accompanying text descriptions. Identify the user's identity type, which includes students, teachers, and parents. Different user identity types correspond to different focuses in the description of needs. For student users, extract the functional expectations related to learning content from their input natural language requirement description; for teacher users, extract the functional expectations related to teaching management from their input natural language requirement description; for parent users, extract the functional expectations related to displaying student learning progress from their input natural language requirement description. Redundant information is filtered out from the extracted user requests of different identity types, and the core request descriptions are retained as the functional page request descriptions entered by users through natural language.

3. The method for generating personalized functional pages based on a large model according to claim 1, characterized in that, The requirement parsing module of the large model performs semantic parsing on the functional page requirement description to obtain the user's functional requirement characteristics and style preference characteristics, including: The functional page requirement description is input into the word vector conversion layer of the large model to generate a corresponding requirement word vector sequence. Each vector in the requirement word vector sequence corresponds to a word unit in the requirement description. The context encoding module of the large model is invoked to perform semantic association modeling on the required word vector sequence, generating encoded vectors containing context dependencies; The encoding vector is split into functional encoding sub-vectors related to functional requirements and style encoding vectors related to style preferences by using the feature separation layer of the large model. The functional encoding subvector is processed by functional dimension mapping to generate functional requirement features that include functional type identifier, interaction logic description and priority order. The functional type identifier is used to distinguish different page function categories, the interaction logic description is used to describe the calling relationship between functions, and the priority order is used to indicate the importance of each function. The style encoding subvector is processed by style dimension mapping to generate style preference features that include color system parameters, font attribute parameters and layout mode parameters. The color system parameters reflect the user's preference for the main color and secondary color of the page, the font attribute parameters reflect the user's preference for font type and size, and the layout mode parameters reflect the user's preference for the arrangement of page elements.

4. The method for generating personalized functional pages based on a large model according to claim 3, characterized in that, The context encoding module of the large model is invoked to perform semantic association modeling on the required word vector sequence, generating encoded vectors containing contextual dependencies, including: The required word vector sequence is input into the multi-head attention layer of the context encoding module, and the attention weight value between each word unit and other word units is calculated. The attention weight value is used to represent the semantic association strength between word units. The required word vector sequence is weighted and aggregated based on the attention weight value to generate a context enhancement vector for each word unit. The context enhancement vector integrates the semantic information of other word units whose relevance to it meets the preset relevance criteria. The context enhancement vector is input into a bidirectional recurrent layer for temporal association modeling, generating a temporal encoding vector containing semantic dependencies between context and context. Each dimension of the temporal encoding vector corresponds to the semantic features of a position in the required word vector sequence. The time-series coding vector is subjected to layer normalization to obtain a standardized time-series coding vector; The standardized temporal encoding vector is transformed by the output layer of the context encoding module to generate an encoding vector with the same length as the required word vector sequence. Each element of the encoding vector contains the contextual semantic information of the corresponding lexical unit.

5. The method for generating personalized functional pages based on a large model according to claim 3, characterized in that, The step of performing functional dimension mapping processing on the functional encoded sub-vectors to generate functional requirement features containing functional type identifiers, interaction logic descriptions, and priority rankings includes: The function classifier is invoked to determine the category of the function encoding sub-vector and output the corresponding function type identifier, which covers information display, operation interaction, data processing and feedback response. Extract the semantic features related to the interaction from the function encoding sub-vector to generate an interaction logic description. The interaction logic description includes the function triggering condition, the function execution flow, and the function result return method. The function triggering condition describes the user operation required to start the function, the function execution flow describes the processing steps inside the function, and the function result return method describes the presentation format of the information after the function is completed. The importance score of each function type is calculated based on the semantic intensity value in the function encoding sub-vector. The semantic intensity value is determined by the projection length of the function encoding sub-vector in the semantic activation direction. The larger the projection length, the higher the degree of emphasis the corresponding function is in the requirement description. Based on the importance score, each function type is sorted to generate a priority sort. Function types that appear earlier in the priority sort are those that should be displayed first when the page is generated. The function type identifier, interaction logic description, and priority sorting are combined according to a preset format to generate the user's functional requirement features.

6. The method for generating personalized functional pages based on a large model according to claim 1, characterized in that, The step of generating an initial functional page layout based on the set of page components, the display attributes of each component, and preset layout rules includes: Extract the function type identifiers and priority order of each component in the page component collection. Based on the functional area division principle in the preset layout rules, divide the page into multiple functional areas, each corresponding to one or more function type identifiers. Based on the priority of each component, the K components with the highest priority in the priority ranking sequence are assigned to functional areas on the page with visual weights that meet the first preset level, and the M components with the lowest priority in the priority ranking sequence are assigned to functional areas on the page with visual weights that meet the second preset level. The visual weights are determined by the position and size of the functional areas on the page, and the visual weights of the first preset level are greater than the visual weights of the second preset level. Based on the size ratio and position parameters in the display attributes of each component, determine the specific placement position and display size of each component in the corresponding functional area so that there is no overlap between components; Analyze the functional relationships between components and, based on the principle of interaction proximity in the preset layout rules, place components with direct interaction relationships in adjacent positions to shorten the user's operation path; The color parameters in the display attributes of each component are globally coordinated to ensure that the color matching of different components conforms to the color system parameters in the style preference characteristics, thus forming a unified page color style. Integrate the placement, display size, color parameters, and functional area division of all components to generate the initial functional page layout.

7. The method for generating personalized functional pages based on a large model according to claim 1, characterized in that, The system monitors user interaction data with the initial functional page in real time, inputs this data into the dynamic adjustment module of the large model, and adjusts the display attributes of page components and the page layout structure based on the interaction data to obtain a personalized functional page, including: Deploy a behavior monitoring component to collect real-time interaction behavior data between users and the initial functional page. The interaction behavior data includes the number of component clicks, the duration of component dwell, the page scrolling trajectory, and the operation sequence. The interactive behavior data is preprocessed to convert the number of component clicks into a click frequency parameter, the component dwell time into a dwell percentage parameter, the page scroll trajectory into a regional attention parameter, and the operation sequence into an interaction path parameter. The preprocessed click frequency parameters, dwell time percentage parameters, regional attention parameters, and interaction path parameters are input into the dynamic adjustment module of the large model to generate user behavior feature vectors. The user behavior feature vector is analyzed by the demand change recognition layer of the dynamic adjustment module to determine whether there is a deviation between the user's current actual needs and the initial functional requirements. If there is a deviation, a demand adjustment coefficient is generated. Based on the aforementioned demand adjustment coefficient, the display attributes of page components are adjusted, increasing the size ratio of components whose click frequency parameter meets the first preset condition, enhancing the color contrast of components whose dwell time parameter meets the second preset condition, and optimizing the component layout of areas whose regional attention parameter meets the third preset condition. The smoothness of user operation is analyzed based on the interaction path parameters. If the operation path length meets the fourth preset condition, the position parameters of the relevant components are adjusted, the page layout structure is optimized, and the interaction path is shortened. The integrated and adjusted page components display attributes and layout structure are used to generate personalized functional pages.

8. The method for generating personalized functional pages based on a large model according to claim 7, characterized in that, The step of analyzing the smoothness of user operations based on interaction path parameters, and if there is a situation where the operation path length meets the fourth preset condition, then adjusting the position parameters of relevant components, optimizing the page layout structure, and shortening the interaction path, includes: Parse the interaction path parameters, extract the sequence of steps the user takes to complete a specific task on the page and the jump trajectory between each step, and generate an operation path diagram; Calculate the jump distance between each step in the operation path diagram. The jump distance is represented by the pixel distance between the center points of the components. Convert the jump distance into a proportional distance relative to the overall page size. Add up the proportional distances between all steps to get the total length of the operation path. The total length of the operation path is compared with a preset path threshold. If the total length of the operation path is greater than the preset path threshold, it is determined that there is an operation path that is too long. Identify the consecutive steps with the longest jump distance in the operation path, and determine the relevant components involved in the consecutive steps; Based on the principle of interaction proximity in the preset layout rules, the position parameters of related components are adjusted to place the components involved in the consecutive steps with the longest jump distance close to each other, thereby reducing the jump distance between them. The total length of the adjusted operation path is recalculated. If the total length of the operation path is still greater than the preset path threshold, the position parameters of the relevant components are repeatedly adjusted until the total length of the operation path is less than or equal to the preset path threshold. Record the adjusted component position parameters and update the page layout structure to optimize the user's operation path.

9. A system for generating personalized functional pages based on large models, characterized in that, The device includes a processor and a memory, the memory and the processor being connected. The memory is used to store programs, instructions or code, and the processor is used to execute the programs, instructions or code in the memory to implement the personalized functional page generation method based on a large model as described in any one of claims 1-8.

Images