Method for replacing business card style based on information synchronization and electronic equipment

Abstract

The invention provides a method for replacing business card styles based on information synchronization and electronic equipment, and relates to the technical field of computers. The method comprises the following steps: receiving an update message of basic information of a current user, and determining a degree score through a pre-trained degree model; inputting the current style definition of the current user, the modified first basic information, the first basic information before modification, the degree score and the definition of the degree score into a generative large model to obtain a plurality of alternative styles; and displaying the plurality of alternative styles so that the current user can select a target style in the alternative styles. And updating and mapping the basic information into a style difference score through a degree model, driving the generative large model to output alternative styles, and completing updating through user selection.

Description

Technical Field

[0001] The present disclosure relates to the field of computer technologies, and more particularly, to a method and an electronic device for changing business card styles based on information synchronization. Background Art

[0002] The replacement of business card styles is usually carried out in the following manner: obtaining the personal information of a user, and generating a QR code image according to the personal information; screening at least one candidate template from a preset template set for the user to select according to the age, gender, and / or position in the personal information; obtaining a selection instruction, and determining the candidate template corresponding to the selection instruction as a personalized template; generating a personalized business card according to the QR code image and the personalized template.

[0003] However, the personalization of this method is limited by the richness of templates. Generally speaking, the number of templates is limited, and the probability of using the same template is relatively high, and there will also be a situation of similarity when using the same template. Summary of the Invention

[0004] In view of this,

[0005] In a first aspect, the present invention provides a method for changing business card styles based on information synchronization, including:

[0006] Receiving an update message of the basic information of the current user, where the update message includes the modified first basic information and the first basic information before modification;

[0007] Based on the modified first basic information and the first basic information before modification, determining a degree score through a pre-trained degree model, where the degree model is used to predict the degree score corresponding to the current update message based on historical update messages, the degree scores corresponding to each historical update message, and the updated style. The degree score is used to indicate the difference size between styles. The style of a business card includes attributes and patterns. The degree score includes an attribute difference score and a pattern difference score. The attribute difference score is used to quantify the impact of the change of text fields in the basic information on the style attributes, and the pattern difference score quantifies the impact of the change of visual fields in the basic information on the style patterns;

[0008] Inputting the current style definition of the current user, the modified first basic information, the first basic information before modification, the degree score, and the definition of the degree score into a generative large model to obtain multiple alternative styles;

[0009] Displaying the multiple alternative styles so that the current user can select a target style from the alternative styles.

[0010] In an alternative embodiment, the attributes include font and layout; the user's basic information includes user personal information, user behavior information, and the scenario information where the user is located. The user personal information includes position information. The first basic information before modification includes the position information before modification, and the first basic information after modification includes the position information after modification.

[0011] In an alternative embodiment, the final degree score takes the maximum value of the attribute difference score and the pattern difference score, and both scores are constrained by the dynamic threshold of the historical style sequence.

[0012] In an alternative embodiment, based on the first basic information after modification and the first basic information before modification, the degree score is determined through a pre-trained degree model, and it further includes:

[0013] Determine the historical target style sequence;

[0014] Based on the historical target style sequence, the first basic information after modification, and the first basic information before modification, the degree score is determined through a pre-trained degree model.

[0015] In an alternative embodiment, the degree model is trained through the following steps:

[0016] Obtain the historical data set: including user basic information update messages, the corresponding historical style change sequence, and the updated target style;

[0017] Extract style features: for each style in the historical style sequence, extract attribute features and pattern features;

[0018] Construct training samples: take the value before modification of each information update, the value after modification of the information update, the corresponding style sequence features, and the manually annotated degree score between the old and new styles as samples;

[0019] Train a dual-path model: the first path learns the attribute difference score; the second path learns the pattern difference score.

[0020] In an alternative embodiment, the dynamic threshold constraint is achieved through the following method:

[0021] Calculate the average attribute difference and average pattern difference of consecutive styles in the historical style sequence;

[0022] If the current attribute difference score > 1.5 times the historical average attribute difference, then suppress it to 60% of the original value;

[0023] If the current pattern difference score > 1.8 times the historical average pattern difference, then suppress it to 40% of the original value.

[0024] In an alternative embodiment, the input of the generative large model further includes:

[0025] Attribute difference score definition: Describing the mapping relationship between the score and the adjustment range of font and layout;

[0026] Pattern difference score definition: Describing the mapping relationship between the score and the adjustment range of color and layout;

[0027] Among them, the large model generates adjustment schemes for the attribute dimension and the pattern dimension respectively according to the two scores, and then combines them into a complete alternative style.

[0028] In an alternative embodiment, the weight allocation of the attribute difference score and the pattern difference score is dynamically adjusted according to the user scenario:

[0029] When the update of basic information involves a business social scenario, increase the weight of the pattern difference score;

[0030] When the update involves a recruitment and job hunting scenario, increase the weight of the attribute difference score.

[0031] In an alternative embodiment, when multiple basic information fields are updated simultaneously:

[0032] Independently calculate the attribute difference score and the pattern difference score for each field;

[0033] When a score conflict is detected, start an arbitration strategy;

[0034] The arbitration strategy includes:

[0035] Give priority to responding to the change with a higher weight of the field, or take the geometric mean of the scores of the conflicting fields as the final value.

[0036] In a second aspect, the present invention provides an electronic device, including:

[0037] One or more processors;

[0038] A memory for storing one or more programs,

[0039] Wherein, when the one or more programs are executed by the one or more processors, the one or more processors are caused to execute the method according to any one of the foregoing embodiments.

[0040] The present invention provides a method and an electronic device for changing business card styles based on information synchronization. By receiving an update message of the basic information of the current user, the update message includes the modified first basic information and the first basic information before modification; based on the modified first basic information and the first basic information before modification, a degree score is determined through a pre-trained degree model, and the degree score includes an attribute difference score and a pattern difference score. The attribute difference score is used to quantify the impact of the change of the text field in the basic information on the style attribute, and the pattern difference score quantifies the impact of the change of the visual field in the basic information on the style pattern; input the current style definition of the current user, the modified first basic information, the first basic information before modification, the degree score, and the definition of the degree score into a generative large model to obtain multiple alternative styles; display the multiple alternative styles so that the current user can select a target style from the alternative styles. By mapping the update of the basic information to the style difference score through the degree model, the generative large model is driven to output alternative styles, and the update is completed after the user's selection. An automated closed-loop from information change to style update is achieved; in the traditional solution, the style needs to be adjusted manually, and the degree model automatically responds to the information change; the degree score quantifies the intensity of the change requirement, avoiding over-design or insufficient adjustment and reducing the user's cognitive load. The user only needs to select from the limited alternative options generated by the AI, and the decision-making time is reduced. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 It is a schematic flowchart of a method for changing business card styles based on information synchronization provided by an embodiment of the present application;

[0042] Figure 2 It is a schematic structural diagram of a device for changing business card styles based on information synchronization provided by an embodiment of the present application;

[0043] Figure 3 It is a schematic structural diagram of an electronic device provided by an embodiment of the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0044] Hereinafter, embodiments of the present disclosure will be described with reference to the accompanying drawings. However, it should be understood that these descriptions are exemplary only and are not intended to limit the scope of the present disclosure. In addition, in the following description, the description of well-known structures and technologies is omitted to avoid unnecessarily confusing the concepts of the present disclosure.

[0045] The terms used herein are for describing specific embodiments only and are not intended to limit the present disclosure. The words "a", "an", and "the" as used herein shall also include the meanings of "multiple" and "plural" unless the context clearly indicates otherwise. In addition, the terms "including", "comprising", etc. as used herein indicate the presence of the described features, steps, operations, and / or components, but do not exclude the presence or addition of one or more other features, steps, operations, or components.

[0046] All terms used herein (including technical and scientific terms) have the meanings commonly understood by those skilled in the art unless otherwise defined. It should be noted that the terms used herein should be interpreted as having a meaning consistent with the context of this specification and should not be interpreted in an idealized or overly rigid manner.

[0047] Figure 1 A schematic flowchart of a method for changing business card styles based on information synchronization provided for the embodiments of this application. As Figure 1 shown, the method includes:

[0048] S110, receiving a current update message of the basic information of the current user. The current update message includes the modified first basic information and the first basic information before modification.

[0049] The style of a business card includes attributes and patterns. The attributes include fonts and layouts; the basic information of a user includes user personal information, user behavior information, and the scene information where the user is located. The user personal information includes position information, and the information that has been updated includes a change in position information.

[0050] The basic information can also generally refer to various basic information related to the user, including personal profiles, behavior habits, and the current environment or scene information.

[0051] The business card style refers to the design form of an electronic or physical business card, including visual elements such as fonts and layouts, and is used to display the user's identity or related information.

[0052] The position information is part of the user personal information and is used to describe the role or title of the user in an organization or scene, such as "manager" or "engineer".

[0053] The update message refers to the notification received by the system about the change in user information, including the specific content before and after modification for comparison and recording.

[0054] The user behavior information refers to the operation records or habit data of the user on a certain system or platform, such as click preferences, browsing history, etc., and is used to analyze the user's behavior patterns.

[0055] S120. Based on the modified first basic information and the first basic information before modification, determine a degree score through a pre-trained degree model, which is used to predict the degree score corresponding to the current update message based on historical update messages, the degree scores corresponding to each historical update message, and the updated style. The degree score is used to indicate the magnitude of the difference between styles. The style of the business card includes attributes and patterns. The degree score includes an attribute difference score and a pattern difference score. The attribute difference score is used to quantify the impact of changes in text fields in the basic information on the style attributes, and the pattern difference score quantifies the impact of changes in visual fields in the basic information on the style patterns.

[0056] Among them, the final degree score takes the maximum value of the attribute difference score and the pattern difference score, and both scores are constrained by the dynamic threshold of the historical style sequence.

[0057] In some embodiments, the input of the generative large model further includes: Attribute difference score definition: describing the mapping relationship between the score and the font / layout adjustment amplitude; Pattern difference score definition: describing the mapping relationship between the score and the color / layout adjustment amplitude; Among them, the large model generates adjustment schemes for the attribute dimension and the pattern dimension respectively according to the two scores, and then combines them into a complete alternative style.

[0058] The degree model is trained through the following steps: Obtain a historical data set: including user basic information update messages and the updated target styles; Extract style features, which may include attribute features and pattern features; Use the values before and after each information update and the manually annotated degree score between the old and new styles as samples; Train a dual-path model: The first path learns the attribute difference score; The second path learns the pattern difference score.

[0059] Among them, the historical data set may include historical update messages, updated styles, and the manually annotated degree scores corresponding to the update messages. The historical data set can be obtained through manual annotation.

[0060] In some embodiments, the weight allocation of the attribute difference score and the pattern difference score is dynamically adjusted according to the user scenario: When the update of the basic information involves a business social scenario, the weight of the pattern difference score is increased; When the update involves a recruitment and job application scenario, the weight of the attribute difference score is increased.

[0061] In some embodiments, when multiple basic information fields are updated simultaneously: Calculate the attribute / pattern difference scores independently for each field; When a score conflict is detected, an arbitration strategy is activated; The arbitration strategy includes: Prioritize responding to changes in fields with higher weights, or take the geometric mean of the scores of the conflicting fields as the final value.

[0062] In some embodiments, the core objective of the degree model can be to receive update messages (pre-modification value, post-modification value) of user basic information and output a numerical "degree score" to quantify the importance level of this information change to the overall visual effect and core information transmission of the business card.

[0063] Purpose of the score: When guiding the generative large language model (LLM) to generate a new style, it indicates the extent of visual adjustment (fine-tuning, moderate adjustment, major reconstruction) that should be made.

[0064] Specific solution of the degree model:

[0065] Model input:

[0066] change_type: The type of information field being changed (string enumeration). For example: "name", "title", "company", "company_logo", "phone", "email", "address", "website", "social_media", "tagline", "headshot", etc.

[0067] change_magnitude: The magnitude of the change, which can be a numerical value or a classification. Calculation method:

[0068] Text fields: Calculate the semantic similarity (using lightweight models such as Sentence-BERT) or edit distance (Levenshtein Distance) between the text before and after the modification. The lower the similarity (or the greater the edit distance), the greater the magnitude.

[0069] Image fields (such as Logo / Headshot): Calculate the perceptual hashing (pHash) difference or structural similarity (SSIM) index between the images before and after the modification. The greater the difference, the greater the magnitude.

[0070] Enumeration or classification fields: Directly determine whether there is a change.

[0071] Numerical fields: Calculate the absolute difference or relative change rate.

[0072] Output: Can be normalized to the range [0, 1], or discretized into {tiny, small, medium, large, huge}.

[0073] field_importance: The preset basic importance weight (numerical value, range such as [0, 10]) of this information field. For example: Company logo: 10; Company: 9; Name: 8; Title: 7; Headshot: 6; Phone / Email: 5; Website / Social media: 4; Address: 3; Slogan: 2.

[0074] user_context (optional, to enhance personalization): The user's historical behavior or attributes.

[0075] user_industry: The industry where the user is located (certain industries are more sensitive to job titles / company names).

[0076] user_role: The user's role (sales vs engineer may have different key points for information display).

[0077] avg_style_change_frequency: How often the user has actively changed the style on average in the past (reflecting their sensitivity to style updates).

[0078] last_change_impact: The degree of style adjustment finally selected by the user after the last modification of the same type of information (if recorded).

[0079] Core of the model and calculation logic (Core Logic):

[0080] Core idea: Degree score = f(Basic importance of field type, Degree of change, [User context]);

[0081] Solution 1: It can be implemented based on a rule-based scoring engine (Rule-based Engine).

[0082] Solution 2: It can be implemented based on a lightweight machine learning model (Light ML Model).

[0083] It is possible to record each information change of the user and record the new style finally selected by the user.

[0084] When making the historical dataset, the degree scores corresponding to each update message can also be generated in the following way, and then the degree model is trained based on the made historical dataset:

[0085] Method 1: Use a computer vision model (such as a pre-trained Siamese network or a network composed of an image feature extractor and a distance calculation module. Compare the business card style picture before the change and the new business card style picture selected by the user, and calculate the visual difference degree (such as the cosine distance of feature vectors, LPIPS perceptual distance). Normalize / map this distance to [0, 10] as the quantization of the actual degree of change.

[0086] Method 2: If the business card style is defined by structured data (such as CSS attributes, template IDs, component configurations), rules can be designed to calculate the differences between the old and new style configurations (such as: Has the layout changed? Has the main color changed? Has the font changed? Have elements been added or deleted?). Combine these differences to obtain an actual degree of change score.

[0087] Definition of target variable: The training target of the degree score is the actually calculated degree of change. Model learning and prediction: Based on this information change, to what extent the user finally actually adjusts the style.

[0088] Training: Use the collected (change features, actual degree of change) data to train the GBDT model.

[0089] Prediction: Input new change features, and the model outputs the predicted degree score (i.e., the predicted degree of adjustment that the user may expect).

[0090] Model output:

[0091] Degree score: A numerical value (e.g., in the range [0, 10]), representing the recommended degree of impact of this information change on the visual style adjustment of the business card. Among them, 0 - 3 can be a minor impact, and it is recommended to make fine-tuning of the style (e.g., update text, fine-tune the position of font size). 4 - 6 is a medium impact, and it is recommended to make medium adjustments (e.g., change the color scheme, adjust the local layout, replace secondary elements). 7 - 10 is a major impact, and it is recommended to make significant reconstruction (e.g., replace the overall layout, visual style, core elements such as the design of the Logo area).

[0092] Explanation of output definition (input to the LLM): Define the score range and the corresponding expected degree of adjustment. For example:

[0093] "The degree score (range: 0 - 10) indicates the amplitude of the visual adjustment recommended for the business card style due to this information change. 0 - 3 points: Minor adjustment (only update content, style remains basically unchanged); 4 - 6 points: Medium adjustment (update content and adjust some visual elements such as color, secondary layout); 7 - 10 points: Major adjustment (it is recommended to explore significantly different layouts, styles, or core visual elements)."

[0094] After deployment, collect data on the actual degree of change. After accumulating sufficient data, train a lightweight ML model (Solution B - GBDT) to replace the rule engine to achieve more accurate and personalized scoring.

[0095] The rule engine is completely independent of historical style data. The ML model depends on the data of the actual adjustment amplitude after the user finally selects the new style, which is much more feasible than annotating "the degree differences between all historical style pairs".

[0096] The rule engine is naturally interpretable. The GBDT model can also provide feature importance, with better interpretability than black-box deep learning models.

[0097] Compatible with LLM input: The output degree score and its clear definition can effectively guide the LLM to control the degree of change in the generated style.

[0098] S130, input the current style definition of the current user, the modified first basic information, the first basic information before modification, the degree score, and the definition of the degree score into the generative large model to obtain multiple alternative styles.

[0099] The current style definition of the current user refers to the style characteristics that the user is currently using or desires. For example, in software interface design, it may include the color scheme of the interface, the layout structure, the font style, etc. It provides the model with the user's basic preferences and requirements for the style.

[0100] After the first basic information is modified, it may change. These changes may be data updates, additions or deletions of attributes, etc. This part of the information reflects the adjusted situation at a certain basic level and provides the model with new basic data basis.

[0101] It is necessary to retain the basic information before modification, which allows the model to understand the initial state before modification, facilitating the model to compare the differences before and after modification, and thus more accurately understand the trend and focus of the changes.

[0102] Definition of the degree score: It clarifies the specific meaning and criteria represented by the degree score, enabling the model to accurately understand the information conveyed by the score, and thus being able to reasonably utilize this score when processing data.

[0103] Inputting this information into the generative large model is to enable the model to comprehensively understand the current situation and background, including the user's preferences, the status of the basic data, and the evaluation of the existing situation, etc. Based on the input information, the generative large model can perform complex data processing, analysis, and generation tasks, thereby obtaining multiple alternative styles. These alternative styles may be various different design schemes, layout methods, style selections, etc., for the user to further select and use.

[0104] For example, the "Style Generation Prompt Template" may include the following content: [Current User Description], [Current Style Definition], [Modified First Basic Information], [First Basic Information before Modification], [Degree Score] Definition of the degree score. Please comprehensively consider all the above information, including the user's current style preference, the modification of the basic information, and the current situation evaluation reflected by the degree score. Based on these input contents, utilize the generative ability to generate [X] alternative styles related to them. These alternative styles should be diverse. On the basis of maintaining a certain rationality and being related to the original style, they can provide differences in different visual effects, layout arrangements, or function presentations, etc. For each generated alternative style, a description as detailed as possible can also be generated correspondingly, including the organization method of interface elements, color application, font style, and other key design elements, so as to clearly understand and evaluate the characteristics and applicability of these styles.

[0105] S140 displays multiple alternative styles so that the current user can select a target style from the alternative styles.

[0106] In some embodiments, the above step S120 can be specifically implemented through the following steps: determining a historical target style sequence; based on the historical target style sequence, the modified first basic information, and the first basic information before modification, determining a degree score through a pre-trained degree model.

[0107] Input of the model that fuses information change and style sequence constraint: Information change: value before modification, value after modification. Style change sequence: an ordered list of the user's historical business card styles (e.g., [Style A, Style B, Style C]), represented as structured features.

[0108] Output: degree score (range 0 - 10), which needs to meet the following: The higher the score: the greater the style adjustment required by the information change (e.g., changing the company logo requires significant adjustment). The lower the score: the information change only requires fine-tuning of the style (e.g., updating the phone number).

[0109] Key constraint: The score needs to refer to the historical style sequence to ensure that the new style and the recent styles will not mutate suddenly (e.g., avoid suddenly changing from a minimalist style to a retro style).

[0110] A dual-path evaluation model can be designed: The model is divided into two paths, which separately process the impact of information change and the coherence of the style sequence, and finally merge the output scores. Among them, the style sequence path (constraining the amplitude of style evolution) can calculate the maximum allowable style change threshold according to the historical style sequence to suppress mutations. Its input features (the style needs to be converted into structured features): Feature type - extraction method - example value; Layout complexity - calculate the distribution dispersion of the main areas (name / company / contact information) - 0.2 (low) → 0.8 (high); Main color system - extract the main peak of the HSV color histogram and encode it as the hue value - [120, 0.7] (green color system); Font style - classification encoding (e.g., 0 = serif, 1 = sans-serif, 2 = handwritten) - 1 (sans-serif); Element density - the proportion of the non-blank area of the business card - 0.6; Proportion of brand elements - the area proportion occupied by brand elements such as logos / slogans - 0.3.

[0111] When the information change requires drastic adjustment but the historical sequence shows that the user's preference for gradual change is low, the degree difference will be significantly suppressed, thus avoiding generating a mutation plan.

[0112] In some embodiments, the dynamic threshold constraint is implemented as follows: calculate the average attribute difference and the average pattern difference between consecutive styles in the historical style sequence; if the current attribute difference score > 1.5 times the historical average attribute difference, suppress it to 60% of the original value; if the current pattern difference score > 1.8 times the historical average pattern difference, suppress it to 40% of the original value.

[0113] Example scenario:

[0114] User A (historical style sequence: minimalist style → micro color matching → new icons added):

[0115] If updating the phone number → info_impact = 2 → final score ≈ 2 (only text update is required);

[0116] If replacing the company logo → info_impact = 9, but the historical max_threshold only allows moderate adjustment → final score = 9 × 0.5 = 4.5 → generate a progressive new style (retain the minimalist framework and only update the logo area);

[0117] User B (frequently switches styles: minimalist → retro → tech):

[0118] Same logo replacement → final score = 9 × 1.0 = 9 → allows significant style reconstruction.

[0119] In the embodiments of this application, the degree model maps the basic information update (before / after values) to a style difference score, drives the generative large model to output alternative styles, and completes the update after user selection. It realizes an automated closed-loop from information change to style update, automatically responds to information changes through the degree model, and improves efficiency.

[0120] The degree score quantifies the intensity of the change requirement (such as phone number modification → low score, company logo replacement → high score). Through constraints, it avoids over-design or insufficient adjustment, reduces the user's cognitive load, and the user only needs to select from the limited alternative options generated by the AI, reducing the decision-making time.

[0121] Establish a structured style description framework, decouple attributes and patterns, lay the foundation for two-dimensional scoring, cover multi-source information-driven scenarios, and in addition to basic information, support behavior data and environmental information to trigger updates.

[0122] The degree model needs to calculate the score in combination with the historical target style sequence to suppress mutations and ensure the coherence of the brand image. Based on the historical style evolution path, it constrains the generation direction of new styles to avoid style jumps.

[0123] By structuring the extraction of historical style features and relying on manually annotated difference scores to train the model, the problem of "difficulty in quantifying style differences" in traditional methods is solved, making the degree score interpretable and significantly improving the generalization ability of the model.

[0124] Figure 2 The following is a schematic structural diagram of a business card style replacement device based on information synchronization provided by an embodiment of the present application. As Figure 2 shown, the device includes:

[0125] A receiving module 201, configured to receive a current update message of the basic information of the current user, where the current update message includes the modified first basic information and the first basic information before modification.

[0126] A scoring module 202, configured to determine a degree score based on the modified first basic information and the first basic information before modification through a pre-trained degree model. The degree model is used to predict the degree score corresponding to the current update message based on historical update messages, the degree scores corresponding to each historical update message, and the updated style. The degree score is used to indicate the difference size between styles. The style of the business card includes attributes and patterns. The degree score includes an attribute difference score and a pattern difference score. The attribute difference score is used to quantify the impact of changes in text fields in the basic information on the style attributes, and the pattern difference score quantifies the impact of changes in visual fields in the basic information on the style patterns.

[0127] A generating module 203, configured to input the current style definition of the current user, the modified first basic information, the first basic information before modification, the degree score, and the definition of the degree score into a generative large model to obtain multiple alternative styles.

[0128] A display module 204, configured to display multiple alternative styles so that the current user can select a target style from the alternative styles.

[0129] In some embodiments, the attributes include fonts and layouts; the basic information of the user includes user personal information, user behavior information, and the scenario information where the user is located. The user personal information includes position information. The first basic information before modification includes the position information before modification, and the modified first basic information includes the position information after modification.

[0130] In some embodiments, the final degree score takes the maximum value of the attribute difference score and the pattern difference score, and both scores are constrained by the dynamic threshold of the historical style sequence.

[0131] In some embodiments, the scoring module 202 specifically includes:

[0132] Determine the historical target style sequence;

[0133] Based on the historical target style sequence, the modified first basic information, and the first basic information before modification, determine the degree score through a pre-trained degree model.

[0134] In some embodiments, the training module is used to obtain a historical data set: including user basic information update messages, corresponding historical style change sequences, and updated target styles; extract style features: for each style in the historical style sequence, extract attribute features and pattern features; construct training samples: use the values before and after each information update, the corresponding style sequence features, and the manually annotated difference scores between the old and new styles as samples; train a dual-path model: the first path learns the attribute difference score; the second path learns the pattern difference score.

[0135] In some embodiments, the dynamic threshold constraint is implemented in the following way:

[0136] Calculate the average attribute difference and average pattern difference of consecutive styles in the historical style sequence;

[0137] If the current attribute difference score > historical average attribute difference × 1.5, suppress it to 60% of the original value;

[0138] If the current pattern difference score > historical average pattern difference × 1.8, suppress it to 40% of the original value.

[0139] In some embodiments, the input of the generative large model further includes:

[0140] Definition of attribute difference score: Describe the mapping relationship between the score and the adjustment range of font and layout;

[0141] Definition of pattern difference score: Describe the mapping relationship between the score and the adjustment range of color and layout;

[0142] Among them, the large model generates adjustment schemes for the attribute dimension and pattern dimension respectively according to the two scores, and then combines them into a complete alternative style.

[0143] In some embodiments, the weight allocation of the attribute difference score and the pattern difference score is dynamically adjusted according to the user scenario:

[0144] When the update of basic information involves the business social scenario, increase the weight of the pattern difference score;

[0145] When the update involves the recruitment and job hunting scenario, increase the weight of the attribute difference score.

[0146] In some embodiments, the scoring module is configured to: independently calculate the attribute and pattern difference scores for each field; initiate an arbitration strategy when a scoring conflict is detected; the arbitration strategy includes: preferentially responding to changes with a higher field weight, or taking the scores of the conflicting fields Figure 3 The geometric mean of the structural schematic diagrams of the electronic devices for implementing the embodiments of the present invention is shown as the final value.

[0147] Such as Figure 3 As shown, the electronic device 300 includes a central processing unit (CPU) 301, which can perform various appropriate actions and processes according to the program stored in the read-only memory (ROM) 302 or the program loaded from the storage section 308 into the random access memory (RAM) 303. In the RAM 303, various programs and data required for the operation of the electronic device 300 are also stored. The CPU 301, ROM 302, and RAM 303 are connected to each other via a bus 304. The input / output (I / O) interface 305 is also connected to the bus 304.

[0148] The following components are connected to the I / O interface 305: an input section 306 including a keyboard, a mouse, etc.; an output section 303 including a cathode ray tube (CRT), a liquid crystal display (LCD), etc. and a speaker, etc.; a storage section 308 including a hard disk, etc.; and a communication section 309 including a network interface card such as a LAN card, a modem, etc. The communication section 309 performs communication processing via a network such as the Internet. A drive 310 is also connected to the I / O interface 305 as needed. A removable medium 311, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., is installed on the drive 310 as needed so that a computer program read from it can be installed into the storage section 308 as needed.

[0149] Specifically, according to the embodiments of the present disclosure, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, the embodiments of the present disclosure include a computer program product including a computer-readable medium carrying instructions. In such an embodiment, the instructions can be downloaded and installed from the network via the communication section 309, and / or installed from the removable medium 311. When the instructions are executed by the central processing unit (CPU) 301, the various method steps described in the present invention are executed.

[0150] Although the example embodiments have been described, it will be apparent to those skilled in the art that various changes and modifications can be made without departing from the spirit and scope of the inventive concept. Therefore, it should be understood that the above example embodiments are not restrictive but illustrative.

Claims

1. A method for changing the business card style based on information synchronization, characterized in that, Including: Receiving a current update message of the basic information of the current user, where the current update message includes the modified first basic information and the first basic information before modification; Based on the modified first basic information and the first basic information before modification, determining a degree score through a pre-trained degree model, where the degree model is used to predict the degree score corresponding to the current update message based on historical update messages, the degree scores corresponding to each historical update message, and the updated style. The degree score is used to indicate the difference size between styles. The style of the business card includes attributes and patterns. The degree score includes an attribute difference score and a pattern difference score. The attribute difference score is used to quantify the impact of changes in text fields in the basic information on the style attributes, and the pattern difference score quantifies the impact of changes in visual fields in the basic information on the style patterns; Inputting the current style definition of the current user, the modified first basic information, the first basic information before modification, the degree score, and the definition of the degree score into a generative large model to obtain multiple alternative styles; Displaying the multiple alternative styles so that the current user can select a target style from the alternative styles.

2. The method according to claim 1, characterized in that, The attributes include font and layout; the basic information of the user includes user personal information, user behavior information, and the scenario information where the user is located. The user personal information includes position information. The first basic information before modification includes the position information before modification, and the modified first basic information includes the position information after modification.

3. The method according to claim 1, wherein The final degree score takes the maximum value of the attribute difference score and the pattern difference score, and both scores are dynamically threshold-constrained by the historical style sequence.

4. The method according to claim 1, characterized in that, Based on the modified first basic information and the first basic information before modification, determining a degree score through a pre-trained degree model further includes: Determining a historical target style sequence; Based on the historical target style sequence, the modified first basic information, and the first basic information before modification, determining a degree score through a pre-trained degree model.

5. The method according to claim 4, wherein The degree model is trained through the following steps: Obtaining a historical data set: including user basic information update messages, corresponding historical style change sequences, and updated target styles; Extracting style features: For each style in the historical style sequence, extracting attribute features and pattern features; Constructing training samples: Using the value before modification of each information update, the value after modification of the information update, the corresponding style sequence features, and the manually annotated degree score between the old and new styles as samples; Training a dual-path model: The first path learns the attribute difference score; the second path learns the pattern difference score.

6. The method according to claim 3, characterized in that, The dynamic threshold constraint is achieved through the following methods: Calculating the average attribute difference and average pattern difference of consecutive styles in the historical style sequence; If the current attribute difference score > 1.5 times the historical average attribute difference, suppressing it to 60% of the original value; If the current pattern difference score > 1.8 times the historical average pattern difference, suppressing it to 40% of the original value.

7. The method according to claim 1, wherein The input of the generative large model further includes: Definition of attribute difference score: Describes the mapping relationship between the score and the adjustment range of font and layout; Definition of pattern difference score: Describes the mapping relationship between the score and the adjustment range of color and layout; Among them, the large model generates adjustment plans for the attribute dimension and the pattern dimension respectively according to the two scores, and then combines them into a complete alternative style.

8. The method according to claim 1, characterized in that The weight allocation of the attribute difference score and the pattern difference score is dynamically adjusted according to the user scenario: When the update of basic information involves the business social scenario, increase the weight of the pattern difference score; When the update involves the recruitment and job hunting scenario, increase the weight of the attribute difference score.

9. The method according to claim 1, wherein When multiple basic information fields are updated simultaneously: Independently calculate the attribute difference score and the pattern difference score for each field; When score conflicts are detected, start the arbitration strategy; The arbitration strategy includes: Prioritize responding to changes with higher field weights, or take the geometric mean of the scores of conflicting fields as the final value.

10. An electronic device, characterized in that, Includes: One or more processors; A memory for storing one or more programs, wherein, when the one or more programs are executed by the one or more processors, the one or more processors are caused to execute the method according to any one of claims 1 to 9.

Images