Display screen layout control method and system based on AI

Abstract

The invention relates to the technical field of display screen layout control, and discloses an AI-based display screen layout control method and system, and the method comprises the steps: obtaining the operation record and focus distribution data of a user, and obtaining the time sequence data of a user behavior; extracting dynamic change rule characteristics of user behaviors according to the time sequence data; user interaction records are mined based on dynamic change rule features, classification is carried out according to time and scene dimensions, and long-term trend features and short-term fluctuation features of user behaviors are determined; obtaining the attention weight value of each layout element, and determining a visual saliency sequence; priority values of high-priority elements are calculated in combination with visual saliency sorting, and a preliminary layout scheme is obtained; performing amplitude adjustment according to the preliminary layout scheme to obtain an adjusted layout scheme, and adaptively changing the sizes and positions of layout elements to obtain an optimized display screen layout. According to the method, the display screen interface layout can be dynamically optimized, and the user interaction experience and the working efficiency are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of display screen layout control, in particular to a display screen layout control method and system based on AI. BACKGROUND

[0002] In the field of modern information interaction, the optimization design of display screen layout is the core link to improve user experience and operation efficiency, which is not only an important direction of human-computer interaction research, but also a key support to promote the landing of intelligent application. Reasonable layout can accurately guide user attention and improve information transmission efficiency, so it plays an irreplaceable role in intelligent terminal, command and control system and other scenes.

[0003] In the prior art, the mainstream display screen layout design method depends on fixed templates or preset rules, that is, a layout scheme is generated by pre-set element size, position and display logic. Although this method can ensure basic usability, it also has significant limitations. On the one hand, fixed templates are difficult to adapt to the operation habits of different users; on the other hand, preset rules cannot respond to the dynamic changes of diversified scenes. More deeply, the existing method lacks the deep use of user behavior data, and the correlation mechanism between user behavior and layout adjustment is not established, resulting in the disconnection between layout design and actual user needs.

[0004] In summary, the existing display screen layout control method cannot dynamically adjust the layout based on user behavior, resulting in low user work efficiency. SUMMARY

[0005] The present application provides a display screen layout control method and system based on AI to dynamically optimize the interface layout and improve user interaction experience and work efficiency.

[0006] In a first aspect, to solve the above technical problems, the present application provides a display screen layout control method based on AI, comprising: Obtaining operation records and focus point distribution data of users in different scenes from a user interaction environment, and storing to obtain time series data stream of user behavior; According to the time series data stream, the operation frequency and focus point migration path of the user in different interaction scenes are extracted to obtain the dynamic change law characteristics of the user behavior; Based on the dynamic change law characteristics, the user interaction records are mined, the user interaction records are classified and arranged according to time dimension and scene dimension to obtain a hierarchical interaction data set, and the long-term trend characteristics and short-term fluctuation characteristics of the user behavior mode are determined according to the hierarchical interaction data set; Determine the layout element set corresponding to different scenes under the pre-established scene classification standard, comprehensively evaluate the long-term trend characteristics and the short-term fluctuation characteristics, obtain the attention weight value of each layout element in the layout element set, and determine the visual saliency ranking of each layout element; Compare the attention weight value with the preset weight threshold to obtain a high-priority element, calculate the priority value of the high-priority element in combination with the visual saliency ranking, and obtain a preliminary layout scheme; According to the preliminary layout scheme, the amplitude adjustment is obtained, and based on the adjusted layout scheme, the size and position of the layout element are adaptively transformed to obtain an optimized display screen layout.

[0007] In an optional implementation, according to the time series data stream, the operation frequency and the attention point migration path of the user in different interactive scenes are extracted to obtain the dynamic change rule characteristics of the user behavior, including: According to the time series data stream, the user behavior data set arranged in time sequence is obtained; From the user behavior data set, the operation frequency and the user attention point containing the time stamp are obtained, if the user attention point changes in adjacent time windows, the start and end positions of the attention point migration path are recorded to obtain the historical trajectory data containing the attention point migration path; When the difference between the operation frequency and the preset average frequency exceeds the preset deviation range, the real-time change state information of the user behavior is extracted; The real-time change state information is compared with the pre-established behavior mode library to obtain a matching degree, when the matching degree exceeds the preset similarity threshold, the dynamic change rule characteristics of the user behavior are determined in combination with the historical trajectory data.

[0008] In an optional implementation, the user interactive record is mined based on the dynamic change rule characteristics, the user interactive record is classified and arranged according to time dimension and scene dimension to obtain a hierarchical interactive data set, including: From the preset user historical behavior database, the complete operation frequency and the attention point migration path associated with the dynamic change rule characteristics are mined to constitute the user interactive record; According to the pre-established time classification rule, the user interactive record is divided into time dimension categories according to different time periods, and is divided into scene dimension categories according to different interactive scenes, and after hierarchical storage, a hierarchical interactive data set is obtained.

[0009] In an optional implementation, the long-term trend characteristics and the short-term fluctuation characteristics of the user behavior mode are determined according to the hierarchical interactive data set, including: According to the layered interaction data set, user operation habit data and interaction frequency characteristics are extracted; By analyzing the repeated behavior characteristics in the user operation habit data, the long-term trend characteristics representing the user behavior patterns are obtained; The interaction frequency characteristics are periodically split, and time periods in which the interaction frequency characteristics exceed a preset frequency threshold are screened out and marked as short-term fluctuation characteristics.

[0010] In an optional implementation, the long-term trend characteristics and the short-term fluctuation characteristics are comprehensively evaluated to obtain attention weight values of each layout element in the layout element set, and a visual saliency ranking of each layout element is determined, including: The long-term trend characteristics and the short-term fluctuation characteristics are hierarchically processed using a pre-established scene classification standard to form a layered behavior frequency distribution data set; The behavior frequency distribution data set is matched with a pre-established behavior frequency-weight mapping table to obtain attention weight values of each layout element by a user; The long-term trend characteristics and the short-term fluctuation characteristics are compared in multiple layers to determine cross-scene repeated behavior frequency distribution characteristics; The attention weight values and the repeated behavior frequency distribution characteristics are combined to comprehensively score each layout element, and the visual saliency ranking of each layout element is determined according to the scores from high to low.

[0011] In an optional implementation, the attention weight values are compared with a preset weight threshold to obtain high-priority elements, and a priority value of the high-priority elements is calculated based on the visual saliency ranking to obtain a preliminary layout scheme, including: Layout elements whose attention weight values exceed a preset weight threshold are marked as high-priority elements; According to the visual saliency ranking, interaction regions corresponding to the top two layout elements are extracted, and in a preset monitoring time window, the gaze dwell time of users in the interaction regions is counted, and regions with gaze dwell time exceeding a preset dwell time threshold are determined as user focus distribution; The user focus distribution and the user operation habit data are fused to obtain a comprehensive feature vector, and a priority value of the high-priority elements is calculated based on the comprehensive feature vector; The priority value corresponds to a proportion adjustment parameter obtained from a pre-established adjustment parameter database; The display position and size proportion of the high-priority elements are optimized according to the proportion adjustment parameter to obtain the preliminary layout scheme.

[0012] In an optional implementation, characterized in that, the step of adjusting the preliminary layout scheme to obtain an adjusted layout scheme, and adaptively transforming the size and position of the layout elements based on the adjusted layout scheme to obtain an optimized display screen layout, comprises: when the proportion adjustment parameter set in the preliminary layout scheme exceeds a preset parameter threshold, limiting the size transformation and position adjustment range of the high-priority elements to obtain an adjusted layout scheme; based on the adjusted layout scheme, performing secondary verification on the size and position of each layout element, and when the verification result shows that part of the layout elements overlap, performing smoothing processing on the overlapping region to obtain final transformation parameters; generating an actual presentation mode of the layout elements according to the final transformation parameters; adjusting the priority value according to the actual presentation mode to adapt to the resolution and proportion limit of the user's display screen to obtain an optimized display screen layout.

[0013] In a second aspect, the present application provides an AI-based display screen layout control system, comprising: an interaction data acquisition module, which acquires operation records and focus point distribution data of a user in different scenarios from a user interaction environment, and stores the data to obtain a time series data stream of user behavior; a behavior rule analysis module, which extracts operation frequency and focus point migration path of a user in different interaction scenarios according to the time series data stream to obtain dynamic change rule characteristics of user behavior; a behavior feature layering module, which mines user interaction records based on the dynamic change rule characteristics, classifies and organizes the user interaction records according to time dimension and scenario dimension to obtain a layered interaction data set, and determines long-term trend characteristics and short-term fluctuation characteristics of user behavior patterns according to the layered interaction data set; an element weight evaluation module, which determines a layout element set corresponding to different scenarios under a pre-established scenario classification standard, comprehensively evaluates the long-term trend characteristics and the short-term fluctuation characteristics, obtains focus weight values of each layout element in the layout element set, and determines a visual saliency ranking of each layout element; a layout decision generation module, which compares the focus weight values with a preset weight threshold to obtain high-priority elements, calculates priority values of the high-priority elements in combination with the visual saliency ranking to obtain a preliminary layout scheme; a layout optimization adjustment module, which adjusts the preliminary layout scheme to obtain an adjusted layout scheme, and adaptively transforms the size and position of the layout elements based on the adjusted layout scheme to obtain an optimized display screen layout.

[0014] Compared with the prior art, the present invention has the following beneficial effects: (1) This invention obtains user operation records and attention distribution data in different scenarios from the user interaction environment and stores them as a time series data stream of user behavior. Then, based on the time series data stream, it extracts the user's operation frequency and attention migration path in different interaction scenarios to obtain the dynamic change pattern characteristics of user behavior. This process can achieve comprehensive and real-time recording of user behavior, dynamically perceive user needs, and enable the system to accurately capture user behavior patterns and preference changes in different scenarios, thereby generating a layout scheme that is more in line with the user's actual operating habits and improving the naturalness and smoothness of user interaction with the display screen.

[0015] (2) This invention mines user interaction records based on dynamic change patterns, classifies and organizes these records according to time and scenario dimensions to obtain a hierarchical interaction data set, and determines the long-term trend characteristics and short-term fluctuation characteristics of user behavior patterns accordingly. In this way, this invention can clearly identify the stability and dynamism of user behavior patterns, more accurately identify user needs in different scenarios, thereby ensuring that the layout achieves the best results in different scenarios and improving the flexibility and adaptability of the layout.

[0016] (3) This invention determines the set of layout elements corresponding to different scenarios under a pre-established scenario classification standard, comprehensively evaluates the long-term trend characteristics and short-term fluctuation characteristics of user behavior patterns, obtains the attention weight value of each layout element, and determines its visual salience ranking, thereby obtaining a preliminary layout scheme. This series of operations can dynamically determine the priority of layout elements, ensure that important information can be displayed to users more prominently, generate a preliminary layout scheme that meets user needs and scenario characteristics, improve information transmission efficiency, and enable users to obtain the information they need more quickly.

[0017] (4) This invention adjusts the size and position of layout elements according to the initial layout scheme to achieve an optimized display layout. This optimization method ensures that layout elements achieve the best display effect in different scenarios, responds to changes in user behavior in real time, and generates a personalized layout that conforms to user operating habits and focus, thereby significantly improving user experience and operating efficiency. Attached Figure Description

[0018] Figure 1 This is a flowchart illustrating an embodiment of the AI-based display layout control method provided by the present invention. Figure 2 This is a schematic diagram of an embodiment of the AI-based display layout control system provided by the present invention. Detailed Implementation

[0019] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by a person of ordinary skill in the art without creative labor fall within the protection scope of the present application.

[0020] With reference to Figure 1 The first embodiment of the present application provides an AI-based display screen layout control method, including steps S11 to S16, specifically as follows: S11, obtaining operation records and attention point distribution data of a user in different scenes from a user interaction environment, and storing to obtain time series data flow of user behavior; S12, extracting operation frequency and attention point migration path of the user in different interaction scenes according to the time series data flow, and obtaining dynamic change rule characteristics of user behavior; S13, mining user interaction records based on the dynamic change rule characteristics, and classifying and arranging the user interaction records according to time dimension and scene dimension to obtain a hierarchical interaction data set; S14, determining long-term trend characteristics and short-term fluctuation characteristics of user behavior mode according to the hierarchical interaction data set; S15, determining a layout element set corresponding to different scenes under a pre-established scene classification standard, comprehensively evaluating the long-term trend characteristics and the short-term fluctuation characteristics, obtaining attention weight values of each layout element in the layout element set, and determining visual saliency ranking of each layout element; S16, comparing the attention weight values with a preset weight threshold to obtain high-priority elements, combining the visual saliency ranking to calculate priority values of the high-priority elements, and obtaining a preliminary layout scheme; S17, performing amplitude adjustment according to the preliminary layout scheme to obtain an adjusted layout scheme, and performing adaptive transformation on sizes and positions of layout elements based on the adjusted layout scheme to obtain an optimized display screen layout.

[0021] In step S11, operation records and attention point distribution data of a user in different scenes are obtained from a user interaction environment, and stored to obtain time series data flow of user behavior.

[0022] It should be noted that the user interaction environment refers to the specific scene and use environment of the user when interacting with the display screen, such as the learning interface of an online education platform, the operation page of a smart office system, etc. Different scenarios correspond to different interaction purposes and behavior patterns of the user. The operation record is the specific action data generated by the user in the interaction process, including the type, object and time of operation such as clicking buttons, sliding pages, and inputting text, which reflects the user's interaction mode and frequency with layout elements. The focus distribution data is the information of the user's gaze focusing area obtained through eye tracking technology or page dwell analysis, which can reflect the user's attention degree to different layout elements. With time as the axis, the operation record and focus distribution data of the user in different scenarios are connected to form a continuous data stream with time sequence markers, i.e. the time series data stream of user behavior.

[0023] Taking the smart office scenario as an example, the user interaction environment is the interface of the office software. The behavior monitoring module built-in the office software will record the user's operation record of clicking the "file transfer" button multiple times from 9:00 to 10:00. At the same time, through page hot area analysis, it is found that the user's gaze stays in the "to-do list" area for the longest time. These data are stored after being marked with the corresponding time stamp to form a time series data stream containing the time sequence relationship of operation and focus, providing raw data for subsequent analysis of the user's office habits.

[0024] In step S12, according to the time series data stream, the operation frequency and focus migration path of the user in different interaction scenarios are extracted, and the dynamic change rule characteristics of the user behavior are obtained, including: According to the time series data stream, the user behavior data set arranged in time sequence is obtained; From the user behavior data set, the operation frequency and user focus containing time stamp are obtained. If the user focus changes within adjacent time windows, the start and end positions of the focus migration path are recorded, and the historical trajectory data containing the focus migration path is obtained; When the difference between the operation frequency and the preset average frequency exceeds the preset deviation range, the real-time change state information of the user behavior is extracted; The real-time change state information is compared with the pre-established behavior pattern library to obtain the matching degree. When the matching degree exceeds the preset similarity threshold, the historical trajectory data is combined to determine the dynamic change rule characteristics of the user behavior.

[0025] It should be noted that the time series data stream is sorted to restore the natural order of user behavior, avoiding analysis deviation caused by data confusion. The operation frequency obtained from the user behavior data set is the number of operations of a user on a specific layout element within a unit time, which can reflect the user's usage dependence on the layout element. The user focus is the interface area where the user's gaze or operation is focused, the adjacent time window refers to a time period divided by a fixed time length, which is continuous in time and does not overlap, and the focus migration path is the trajectory of the focus switching from an interactive area corresponding to one layout element to an interactive area corresponding to another layout element in adjacent time windows. These trajectory data collectively constitute historical trajectory data.

[0026] In the embodiment, the preset average frequency is a regular operation frequency calculated based on historical data, and the preset deviation range is a normal fluctuation interval allowed. When the difference between the operation frequency and the preset average frequency exceeds the preset deviation range, it indicates that the user behavior has abnormal fluctuations, and information extraction is performed on the operation object, frequency, time, etc. of the abnormal fluctuations to obtain real-time change state information. This process can capture temporary changes in user behavior and avoid ignoring short-term special needs by only performing long-term trend analysis.

[0027] Further, the real-time change state information is disassembled into a plurality of real-time feature parameters, including the operation object, the operation frequency, the time period, the area where the focus is located, etc. The behavior pattern library established in advance stores the feature parameters of various typical behavior patterns. The coincidence degree or the degree of similarity between the real-time feature parameters and the feature parameters of each pattern in the behavior pattern library is calculated, for example, the consistent proportion of the two in the operation object, the frequency interval, the time period distribution, etc. dimensions are calculated to obtain the matching degree. When the matching degree exceeds the preset similarity threshold, the records related to the current behavior in the historical trajectory data are retrieved, and the consistency of the current behavior characteristics and the historical behavior characteristics is analyzed, such as the fluctuation trend of the operation frequency, the similarity of the focus migration path, etc. to form dynamic change rule characteristics with both short-term fluctuations and long-term trends.

[0028] The setting of the preset similarity threshold needs to consider the matching accuracy requirement of the application scenario and refer to the effective matching feature distribution in the historical data. If the matching result is required to have high accuracy, the threshold can be set to a high level (such as 70%-80%); if potential behavior changes need to be captured preferentially, the threshold can be appropriately reduced (such as 60%-70%).

[0029] Exemplarily, in the intelligent office scene, taking 15 minutes as a time window, after the time series data stream is sorted by timestamp, the user's behavior from 14:00 to 14:30 is presented. In the period from 14:00 to 14:30, the operation frequency of the "to-do list" is 4 times / hour, and the click frequency of the "file transfer" button is 8 times / hour. In the period from 14:00 to 14:15, the user focuses on the interaction area corresponding to the "to-do list", and in the period from 14:15 to 14:30, the user switches to the interaction area corresponding to the "file transfer" button. Thus, the user's attention migration path is recorded, constituting the historical trajectory data. The preset average frequency of file transfer is 2 times / hour, and the deviation range is ±1 time / hour. The difference between the operation frequency of the "file transfer" button by the user in the period from 14:00 to 14:30 and the preset average frequency exceeds the preset deviation range, and the real-time change state information is extracted.

[0030] In the present embodiment, the behavior pattern library established in advance stores a typical behavior pattern of "high-frequency use of file transfer function in project promotion stage", and the characteristic parameters thereof include operation object "file transfer" button, operation frequency 4-8 times / hour, and occurrence period mostly in the afternoon, etc. The characteristic parameters of the real-time change state information are compared with the characteristic parameters of the mode, and it is calculated that the matching degree is 85%, which exceeds the preset similarity threshold of 70%. In combination with similar behaviors in the historical trajectory in the past week, it is determined that the dynamic change rule feature of the user behavior is "in the period from 14:00 to 14:30 in the project promotion stage, the user will operate the "file transfer" button with high frequency, and the attention point will migrate from the "to-do list" to the "file transfer" button".

[0031] In step S13, based on the dynamic change rule feature, the user interaction record is mined, and the user interaction record is classified and arranged according to time dimension and scene dimension to obtain a hierarchical interaction data set, including: From the preset user historical behavior database, the complete operation frequency and attention point migration path associated with the dynamic change rule feature are mined to constitute the user interaction record; According to the pre-established time classification rule, the user interaction record is divided into different time periods and classified into time dimension categories, and at the same time, it is divided into different interaction scenes and classified into scene dimension categories, and after hierarchical storage, a hierarchical interaction data set is obtained.

[0032] It should be noted that the preset user historical behavior database is a collection of storing all past interaction behavior data of the user, covering operation and focus information in different periods and different scenarios. The dynamic change rule feature includes core attributes of user behavior, such as specific operation object, operation frequency range, focus migration direction, and associated scenario, and the core attributes are used as retrieval basis to perform data filtering in the user historical behavior database to extract complete operation frequency and focus migration path that match the dynamic change rule feature. The completeness of the operation frequency is reflected in the operation times and time distribution within a specific period, and the completeness of the focus migration path is reflected in the complete start and end area and time record. After the operation frequency data and the focus migration path data filtered and verified for completeness are aggregated, the user interaction record is formed.

[0033] It should be further noted that the time classification rule is a division standard for time periods, such as division according to hours, weekdays / weekends, months, etc., and its purpose is to reflect the distribution characteristics of user behavior in time, such as the difference in use of a function in the morning and in the afternoon; the scenario dimension category is a scenario type divided according to interaction purposes, such as project advancement and daily communication in an office scenario, and course learning and exercise in an education scenario, which is used to distinguish the behavior differences under different purposes. The hierarchical storage is to classify and save the interaction records in the same time and the same scenario to obtain a structured form of the hierarchical interaction data set. For example, the file transfer operation records in the project advancement scenario from 9:00 to 10:00 in the morning on weekdays are stored separately to avoid mixing of data in different times and scenarios, and to facilitate subsequent quick extraction of data analysis of user behavior patterns in specific dimensions.

[0034] In step S14, long-term trend characteristics and short-term fluctuation characteristics of a user behavior pattern are determined according to the hierarchical interaction data set, including: According to the hierarchical interaction data set, user operation habit data and interaction frequency characteristics are extracted; By analyzing the repeated behavior characteristics in the user operation habit data, the long-term trend characteristics of the user behavior pattern are obtained; The interaction frequency characteristics are periodically split, and time periods in which the interaction frequency characteristics exceed a preset frequency threshold are filtered out and marked as the short-term fluctuation characteristics.

[0035] It should be noted that the operation habit data is extracted from the hierarchical interaction data set, which is an operation preference with a regularity, such as a fixed operation sequence in which the user tends to view to-do items before file transfer during project advancement. The interaction frequency characteristics are distribution characteristics of the operation times in the time dimension, including the operation times per unit time and the operation intensive period. The operation habit data reflects the stability of user behavior, and the interaction frequency characteristics reflect the dynamic change of user behavior.

[0036] In this embodiment, the repeated behavior feature refers to consistent behavior of the user appearing multiple times in different periods and the same scene. By analyzing the repeated behavior feature in the user operation habit data, the core demand habit of the user can be summarized, the long-term trend characteristics are obtained, the stability and regularity of the user behavior are embodied, and the visibility needs to be considered first during layout. At the same time, the periodic splitting of the interaction frequency feature is to decompose the interaction frequency according to the time period (such as day, week, month), and to distinguish the stable frequency and the sudden frequency change in the regular period. The preset frequency threshold is the upper limit of the normal frequency fluctuation set based on the long-term trend characteristics. When the interaction frequency of a certain time period exceeds the preset frequency threshold, it indicates that the behavior deviates from the regularity at this time, and the time period is marked as a short-term fluctuation characteristic. The short-term fluctuation characteristic reflects the temporary demand of the user in the special scene, and can ensure that the layout can flexibly adapt to the sudden demand.

[0037] Exemplarily, in the hierarchical interaction data set of the intelligent office scene, the operation data of the user in the past half year project promotion scene is extracted. From 9:00 to 11:00 every Monday to Friday in the morning, the operation frequency of “file transmission” is stable at 3 times / hour, and the operation sequence is always from viewing “to-do list” to clicking “file transmission”. This repeated behavior feature is determined as a long-term trend characteristic; while in the morning of a Friday (project deadline) in a week, the operation frequency of “file transmission” increases to 7 times / hour, which exceeds the preset frequency threshold of 5 times / hour, so this time period is marked as a short-term fluctuation characteristic, reflecting the user's urgent transmission demand at the project deadline.

[0038] In step S15, the set of layout elements corresponding to different scenes is determined under the pre-established scene classification standard, the long-term trend characteristics and the short-term fluctuation characteristics are comprehensively evaluated, the attention weight value of each layout element in the set of layout elements is obtained, and the visual saliency ranking of each layout element is determined, including: The long-term trend characteristics and the short-term fluctuation characteristics are hierarchically processed by using the pre-established scene classification standard, to form a hierarchical behavior frequency distribution data set; The behavior frequency distribution data set is matched with the pre-established behavior frequency-weight mapping table to obtain the attention weight value of each layout element for the user; The long-term trend characteristics and the short-term fluctuation characteristics are compared in multiple layers to determine the repeated behavior frequency distribution characteristics across scenes; The attention weight value and the repeated behavior frequency distribution characteristics are combined to comprehensively score each layout element, and the visual saliency ranking of each layout element is determined according to the score from high to low.

[0039] It should be noted that the pre-established scene classification standard is a scene division rule formulated based on user interaction purposes, and each scene corresponds to a specific set of layout elements, such as file processing, schedule management, and other elements in the office scene, and playback control, content recommendation, and other elements in the entertainment scene. Its role is to clarify the core interactive objects in different scenes. Hierarchical processing is to classify long-term trend characteristics and short-term fluctuation characteristics by scene, so that each characteristic can correspond to the layout elements of a specific scene. The behavior frequency distribution dataset formed will record the long-term average frequency and short-term peak frequency of elements by scene, providing a scene-based quantitative basis for subsequent weight calculation. The behavior frequency-weight mapping table is a corresponding rule between frequency and weight obtained by training a large amount of historical data, which contains weight values corresponding to different operation frequencies. Its role is to convert abstract frequency data into quantifiable weight indicators. The matching process is to extract the behavior frequency data of elements by scene, and determine the attention weight value of each layout element according to the behavior frequency-weight mapping table, which can directly reflect the demand intensity of users for each layout element in the corresponding scene.

[0040] In this embodiment, the multi-layer comparison of long-term trend characteristics and short-term fluctuation characteristics will be carried out from the dimensions of behavior consistency and element association, etc. For example, whether the behavior frequency of the same element in daily office and emergency office is higher than that of other elements, or whether the high-frequency elements in different scenes have functional association, such as file transmission and cloud storage are often used simultaneously at high frequency. The repeated behavior frequency distribution characteristics across scenes refer to stable interaction rules that continuously appear in multiple scenes, which reflect the universality value of elements in the user's overall interaction logic. Even if a certain element does not have a prominent weight in a single scene, but it repeatedly and stably appears in multiple scenes, it also shows that it has a basic and important role for the user.

[0041] Further, the attention weight value is taken as the basic item, and the proportion is set to 70%, that is, the attention weight value is multiplied by 70 points to get the basic points, and the repeated behavior frequency distribution characteristics across scenes are taken as additional points, and each additional scene adds 5 points (limited to a maximum of 6 repeated scenes), forming a comprehensive scoring mechanism. According to the comprehensive score from high to low, the interface layout elements are sorted to obtain the visual saliency ranking, and its core role is to enable users to quickly locate the most needed elements during interaction, reduce the operation path length, and improve the interaction efficiency.

[0042] By way of example, the scene classification criteria are divided into three scenes of contract drafting, cross-department collaboration and emergency approval according to the interaction purposes, which correspond to layout elements such as "text input", "online meeting" and "approval submission". In hierarchical processing, long-term trend characteristics and short-term fluctuation characteristics are classified by scene. The behavior frequency distribution dataset records that the long-term average operation frequency of "text input" in the contract drafting scene is 8 times / hour, the long-term average operation frequency of "online meeting" in the cross-department collaboration scene is 5 times / hour, and the short-term peak frequency of "approval submission" in the emergency approval scene is 6 times / hour. For the behavior frequency-weight mapping table, after matching the element frequency data by scene, the attention weight values corresponding to "text input", "online meeting" and "approval submission" are 0.7, 0.5 and 0.8 respectively.

[0043] Further, after multi-layer comparison, it is found that "online meeting" is not only used stably in the cross-department collaboration scene, but also repeatedly used in the contract drafting scene to communicate details with colleagues; "text input" is only used stably in the contract drafting scene; and "approval submission" is only used frequently in the emergency approval scene. According to the comprehensive scoring mechanism, the basic score of "text input" is 0.7x70=49 points, no additional points, and the comprehensive score is 49 points; the basic score of "online meeting" is 0.5x70=35 points, and an additional 5 points are added because there is one more high-frequency scene, and the comprehensive score is 40 points; the basic score of "approval submission" is 0.8x70=56 points, and no additional points are added, and the comprehensive score is 56 points. Therefore, the final visual saliency ranking is: approval submission-text input-online meeting.

[0044] In step S16, the attention weight value is compared with the preset weight threshold to obtain a high-priority element, and the priority value of the high-priority element is calculated based on the visual saliency ranking to obtain a preliminary layout scheme, including: marking the layout element with an attention weight value exceeding the preset weight threshold as a high-priority element; According to the visual saliency ranking, the layout elements corresponding to the top two ranked interactive areas are extracted, and the gaze dwell time of the user in the interactive area is counted within a preset monitoring time window. The area with a gaze dwell time exceeding a preset dwell time threshold is determined as a user focus distribution. Multi-dimensional feature fusion is performed on the user focus distribution and the user operation habit data to obtain a comprehensive feature vector, and the priority value of the high-priority element is calculated based on the comprehensive feature vector; obtaining the proportion adjustment parameter corresponding to the priority value from a pre-established adjustment parameter database; According to the proportion adjustment parameter, the display position and size proportion of the high-priority element are optimized to obtain the preliminary layout scheme.

[0045] It should be noted that the preset weight threshold is a core element screening standard set based on user interaction needs, and layout elements with attention weight values exceeding the preset weight threshold are marked as high-priority elements, which can avoid non-core elements occupying high-quality layout resources. The top two elements in the visual saliency ranking are the objects that the user pays the most attention to, and monitoring the gaze dwell time of the interaction area of the top two elements can verify whether the element position conforms to the user's natural gaze rules. The distribution state formed by the interaction area with a user gaze dwell time exceeding a preset dwell time threshold is the user focus distribution. From the user focus distribution, the position of the focus area (such as the left side, center, etc. of the interface) and the proportion of the gaze dwell time within the monitoring window are extracted, and features such as operation frequency and operation sequence are extracted from user operation habit data. These features are converted into numerical vectors of a unified dimension, such as converting the focus area position into a coordinate value, converting the gaze dwell time into a time proportion within the monitoring window, dividing the operation frequency by the standard frequency, and converting the operation sequence into a sequence value according to the preset step weight. Through multi-dimensional feature fusion processing, the originally dispersed visual features and operation features are scaled to values between 0 and 1, and after integration, a comprehensive feature vector containing multi-dimensional information such as position, time proportion, operation frequency, and operation sequence is obtained.

[0046] For example, in an online shopping scenario, "product search" is located in the upper-central area of the interface, which is quantified as 0.8 according to the preset coordinate rule (1.0 at the top of the interface and 0 at the bottom), so the focus area position corresponds to the value 0.8; within a 5-minute monitoring window, the user's gaze dwells in the focus area for 3.5 minutes, and 3.5 divided by 5 gives 0.7, so the gaze dwell time proportion value is 0.7; since the user clicks "product search" 9 times within 1 hour, the operation frequency feature value is 0.9 after dividing the operation frequency by the standard frequency; after entering the interface, the user first browses the product (step 1) and then clicks search (step 2), and the operation sequence value is determined to be 0.6 according to the preset operation step weight proportion. The final independent quantitative data is integrated into a comprehensive feature vector containing focus area position, gaze dwell proportion, operation frequency, and operation sequence (0.8, 0.7, 0.9, 0.6).

[0047] Different features have different importance in priority evaluation, and each dimension feature in the vector needs to be weighted and summed through a preset weight allocation rule. Specifically, operation frequency and gaze dwell time proportion are core features, and the weights can be set to 0.3 respectively; focus area position and operation sequence are auxiliary features, and the weights can be set to 0.2 respectively. When calculating, the values of each dimension in the comprehensive feature vector are multiplied by the corresponding weights and summed, and the result is the priority value. Taking the comprehensive feature vector of "product search" (0.8, 0.7, 0.9, 0.6) as an example, according to the above weight proportion, the priority value of "product search" can be calculated as 0.76.

[0048] In the embodiment, the adjustment parameter database stores layout optimization rules corresponding to different priority values, including display positions and size proportions of layout elements. After obtaining the priority value of the high-priority element, the corresponding display position and size proportion rule is matched from the adjustment parameter database, and the element layout is adjusted according to the rule to obtain a preliminary layout scheme, thereby improving the efficiency of user searching and operation. Specifically, when the priority value of the "commodity search" is 0.76, the corresponding layout rule is matched from the adjustment parameter database, which is to be displayed in the left 20% and upper 10% area of the interface, and the horizontal and vertical size proportions are both 28%. After adjustment, the preliminary scheme of the element layout is formed.

[0049] In step S17, an adjusted layout scheme is obtained according to the preliminary layout scheme, and the size and position of the layout element are adaptively transformed based on the adjusted layout scheme to obtain an optimized display screen layout, including: When the proportion adjustment parameter set in the preliminary layout scheme exceeds the preset parameter threshold, the size transformation and position adjustment range of the high-priority element are limited to obtain an adjusted layout scheme; Based on the adjusted layout scheme, the size and position of each layout element are secondarily checked, and when the checking result shows that part of the layout elements overlap, the overlapping area is smoothed to obtain final transformation parameters; According to the final transformation parameters, the actual presentation mode of the layout element is generated; According to the actual presentation mode, the priority value is adjusted to adapt to the resolution and proportion limit of the user's display screen to obtain an optimized display screen layout.

[0050] It should be noted that the preset parameter threshold is an upper limit of the position offset and size proportion set to avoid excessive occupation of the interface space by the element, and is used to maintain the visual balance of the interface elements. For example, the preset parameter threshold limits the size proportion to be at most 25% and the position offset from the boundary to be at least 5%, and the preliminary size proportion of the "commodity search" is 28%, which exceeds the size threshold of 25%. Therefore, it is limited to 25% to prevent other elements from being excessively squeezed.

[0051] The secondary verification is implemented by assigning an independent coordinate interval to each layout element and importing the coordinate intervals of all elements into the virtual interface model. Whether the elements overlap is determined by calculating the overlapping area ratio. When the coordinate interval overlapping area exceeds 10% of the area of a single element, the verification result is determined as element overlap. The smoothing processing of the overlapping area needs to be combined with the element priority and the interface space margin. If a high-priority element overlaps with a low-priority element, the position of the low-priority element is adjusted first. If the priorities of the overlapping elements are similar, the size of one of the elements is reduced while keeping the core operation area intact until the overlapping area is less than 10%. The final transformation parameters are the element size and position values determined after smoothing processing, which are converted from the coordinate intervals and size ratios after the secondary verification.

[0052] For example, in an e-commerce APP interface, the "product search box" and the "coupon pop-up window" are two layout elements. According to the priority value, the "product search box" is assigned a coordinate interval of 20%-50% horizontally and 10%-20% vertically, and the "coupon pop-up window" is assigned a coordinate interval of 35%-65% horizontally and 15%-25% vertically. After importing these two coordinate intervals into the virtual interface model, it is calculated that the overlapping area is 25% of the area of the "product search box", which exceeds the threshold of 10%, and the verification result determines that the elements overlap. Since the priority of the "product search box" is higher than that of the "coupon pop-up window", and the preset minimum position offset from the boundary is 5%, there is a 30% space margin on the right side of the interface, the position of the "coupon pop-up window" is adjusted, and its horizontal coordinate interval is offset to the right by 10%, becoming 45%-75%, and the vertical coordinate interval remains unchanged. The overlapping area is calculated again, and at this time the overlapping area is only 8.3% of the area of the "product search box", which is lower than the threshold of 10%. After smoothing processing, the coordinate interval of the "coupon pop-up window" is stable at 45%-75% horizontally and 15%-25% vertically, and the final transformation parameters converted from this are 45% and 15% in horizontal and vertical positions, and 30% and 10% in horizontal and vertical size ratios.

[0053] According to the final transformation parameters, the elements are displayed in the interface according to the determined size and position, generating the actual presentation mode of the layout elements. To adapt to the resolution and proportion limit of the user's display screen, the screen parameters need to be obtained first, and then the actual display size of the elements is converted according to the pixel density. For example, for a 1080P resolution corresponding to 1920x1080 pixels and a 16:9 display ratio, a 20% size corresponds to 216 pixels. If the converted size is less than 100 pixels, the display area needs to be re-assigned, and the feature weight value of the focus area position is increased to improve the element priority value, ensuring that the element is still clear and visible on a low-resolution screen. If the screen ratio is 4:3 instead of 16:9, the horizontal size needs to be compressed and the vertical size needs to be expanded according to the ratio to avoid the element exceeding the screen boundary, and finally an optimized display screen layout is formed.

[0054] In summary, the application discloses an AI-based display screen layout control method, comprising: obtaining user operation records and focus point distribution data in different scenes from a user interaction environment, and storing the data to obtain time series data flow of user behavior; extracting operation frequency and focus point migration path of the user in different interaction scenes according to the time series data flow, and obtaining dynamic change rule characteristics of user behavior; mining user interaction records based on the dynamic change rule characteristics, classifying and arranging the user interaction records according to time dimension and scene dimension to obtain a hierarchical interaction data set, and determining long-term trend characteristics and short-term fluctuation characteristics of user behavior patterns according to the hierarchical interaction data set; determining a layout element set corresponding to different scenes under a pre-established scene classification standard, comprehensively evaluating the long-term trend characteristics and the short-term fluctuation characteristics, obtaining focus weight values of each layout element in the layout element set, and determining visual saliency ranking of each layout element; comparing the focus weight values with a preset weight threshold to obtain high-priority elements, calculating priority values of the high-priority elements in combination with the visual saliency ranking, obtaining a preliminary layout scheme; performing amplitude adjustment according to the preliminary layout scheme to obtain an adjusted layout scheme, and adaptively transforming sizes and positions of layout elements based on the adjusted layout scheme to obtain an optimized display screen layout. The application accurately captures the interaction needs of the user in different scenes through deep mining and multi-dimensional analysis of user behavior data, realizes intelligent adaptation of the display screen layout according to user behavior habits, and improves the convenience and efficiency of user interaction.

[0055] Reference Figure 2 The second embodiment of the application provides an AI-based display screen layout control system, comprising: An interaction data acquisition module obtains user operation records and focus point distribution data in different scenes from a user interaction environment, and stores the data to obtain time series data flow of user behavior; A behavior rule analysis module extracts operation frequency and focus point migration path of the user in different interaction scenes according to the time series data flow, and obtains dynamic change rule characteristics of user behavior; A behavior feature hierarchical module mines user interaction records based on the dynamic change rule characteristics, classifies and arranges the user interaction records according to time dimension and scene dimension to obtain a hierarchical interaction data set, and determines long-term trend characteristics and short-term fluctuation characteristics of user behavior patterns according to the hierarchical interaction data set; An element weight evaluation module determines a set of layout elements corresponding to different scenes under a pre-established scene classification standard, comprehensively evaluates the long-term trend characteristics and the short-term fluctuation characteristics, obtains an attention weight value of each layout element in the set of layout elements, and determines a visual saliency ranking of each layout element; A layout decision generation module compares the attention weight value with a preset weight threshold to obtain a high-priority element, calculates a priority value of the high-priority element in combination with the visual saliency ranking, and obtains a preliminary layout scheme. A layout optimization adjustment module performs amplitude adjustment on the preliminary layout scheme to obtain an adjusted layout scheme, and performs adaptive transformation on the size and position of the layout elements based on the adjusted layout scheme to obtain an optimized display screen layout.

[0056] It should be noted that the AI-based display screen layout control system provided by the embodiments of the present application is used to execute all process steps of the AI-based display screen layout control method provided by the above embodiments, and the working principles and beneficial effects of the two are one-to-one correspondence, so they will not be repeated here.

[0057] The embodiments of the present application also provide an electronic device. The electronic device includes a processor, a memory, and a computer program, such as a behavior pattern matching program, stored in the memory and executable on the processor. The processor implements the steps in the above various AI-based display screen layout control method embodiments when executing the computer program, such as Figure 1 The steps S11 shown. Alternatively, the processor implements the functions of each module / unit in the above various device embodiments when executing the computer program, such as the interactive data acquisition module.

[0058] For example, the computer program can be divided into one or more modules / units, which are stored in the memory and executed by the processor to complete the present application. The one or more modules / units can be a series of computer program instruction segments that can complete a specific function, which are used to describe the execution process of the computer program in the electronic device.

[0059] The electronic device can be a desktop computer, a notebook, a palm computer, and a smart tablet, etc. The electronic device can include, but is not limited to, a processor, a memory. Those skilled in the art can understand that the above components are only examples of the electronic device and do not constitute a limitation on the electronic device, and can include more or fewer components than the above, or combine certain components, or different components, for example, the electronic device can also include an input / output device, a network access device, a bus, etc.

[0060] The processor can be a central processing unit (CPU), and can also be other general-purpose processors, a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field-programmable gate array (FPGA) or other programmable logic device, discrete gate or transistor logic, discrete hardware components, etc. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor. The processor is a control center of the electronic device, and connects various parts of the electronic device through various interfaces and lines.

[0061] The memory can be used to store the computer program and / or modules, and the processor realizes various functions of the electronic device by running or executing the computer program and / or modules stored in the memory, and calling data stored in the memory. The memory can mainly include a program storage area and a data storage area. The program storage area can store an operating system, at least one application program required by a function (such as a sound playing function, an image playing function, etc.), etc. The data storage area can store data created according to the use of the mobile phone (such as audio data, a phone book, etc.), etc. In addition, the memory can include a high-speed random access memory, and can also include a nonvolatile memory, such as a hard disk, a memory, a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, at least one disk storage device, a flash memory device, or other volatile solid-state memory device.

[0062] The modules / units integrated in the electronic device, if realized in the form of software function units and sold or used as independent products, can be stored in a computer readable storage medium. Based on such understanding, all or part of the processes in the above-mentioned embodiment methods can also be completed by a computer program instructing related hardware, and the computer program can be stored in a computer readable storage medium. The computer program can implement the steps of each method embodiment when executed by a processor. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or some intermediate forms, etc. The computer readable medium can include any entity or device, recording medium, U disk, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signal, telecommunication signal, and software distribution medium, etc. that can carry the computer program code. It should be noted that the contents included in the computer readable medium can be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction, for example, in some jurisdictions, according to legislation and patent practice, the computer readable medium does not include electrical carrier signals and telecommunication signals.

[0063] It should be noted that the above-described device embodiments are only schematic, and the units described as separate components can or can not be physically separated, and the components shown as units can or can not be physical units, that is, they can be located in one place, or distributed on multiple network units. Part or all of the modules can be selected according to actual needs to achieve the purpose of the embodiment. In addition, the connection relationship between the modules in the device embodiment provided by the present application indicates that there is a communication connection between them, which can be realized as one or more communication buses or signal lines. Those skilled in the art can understand and implement it without creative labor.

[0064] The above-described specific embodiments further illustrate the purpose, technical solutions and beneficial effects of the present application. It should be understood that the above-described specific embodiments are only for the specific embodiments of the present application and do not limit the protection scope of the present application. It is particularly pointed out that any modification, equivalent replacement, improvement, etc. made by those skilled in the art within the spirit and principles of the present application should be included in the protection scope of the present application.

Claims

1. An AI-based display screen layout control method, characterized by, The application comprises the following steps: Obtain the operation records and focus point distribution data of users in different scenarios from the user interaction environment, and store the time series data stream of user behavior after storage; According to the time series data stream, extract the operation frequency and focus point migration path of users in different interaction scenarios, and obtain the dynamic change rule characteristics of user behavior; Based on the dynamic change rule characteristics, mine the user interaction records, classify and arrange the user interaction records according to the time dimension and scene dimension to obtain a hierarchical interaction data set, and determine the long-term trend characteristics and short-term fluctuation characteristics of the user behavior mode according to the hierarchical interaction data set; Determine the layout element set corresponding to different scenes under the pre-established scene classification standard, comprehensively evaluate the long-term trend characteristics and short-term fluctuation characteristics, obtain the focus weight value of each layout element in the layout element set, and determine the visual saliency ranking of each layout element; Compare the focus weight value with the preset weight threshold to obtain high-priority elements, calculate the priority value of the high-priority elements combined with the visual saliency ranking, and obtain a preliminary layout scheme; According to the preliminary layout scheme, make amplitude adjustment to obtain an adjusted layout scheme, and based on the adjusted layout scheme, adaptively transform the size and position of the layout elements to obtain an optimized display screen layout. 2.The AI-based display screen layout control method of claim 1, wherein, According to the time series data stream, extract the operation frequency and focus point migration path of users in different interaction scenarios, and obtain the dynamic change rule characteristics of user behavior, including: Sort the time series data stream to obtain a user behavior data set arranged in time sequence; From the user behavior data set, obtain the operation frequency and user focus point containing time stamp, if the user focus point changes in adjacent time window, record the start and end position of the focus point migration path, and obtain historical trajectory data containing the focus point migration path; When the difference between the operation frequency and the preset average frequency exceeds the preset deviation range, extract the real-time change state information of user behavior; Compare the real-time change state information with the pre-established behavior mode library to obtain the matching degree, and when the matching degree exceeds the preset similarity threshold, combine the historical trajectory data to determine the dynamic change rule characteristics of user behavior. 3.The AI-based display screen layout control method of claim 2, wherein, Based on the dynamic change rule characteristics, mine the user interaction records, classify and arrange the user interaction records according to the time dimension and scene dimension to obtain a hierarchical interaction data set, including: From the preset user historical behavior database, mine the complete operation frequency and focus point migration path associated with the dynamic change rule characteristics to constitute the user interaction records; According to the pre-established time classification rule, divide the user interaction records into time dimension categories according to different time periods, and into scene dimension categories according to different interaction scenarios, and store them hierarchically to obtain a hierarchical interaction data set. 4.The AI-based display screen layout control method of claim 3, wherein, According to the hierarchical interaction data set, determine the long-term trend characteristics and short-term fluctuation characteristics of the user behavior mode, including: According to the hierarchical interaction data set, user operation habit data and interaction frequency characteristics are extracted; By analyzing the repeated behavior characteristics in the user operation habit data, the long-term trend characteristics representing the user behavior mode are obtained; The interaction frequency characteristics are periodically split, and time periods in which the interaction frequency characteristics exceed a preset frequency threshold are screened out and marked as short-term fluctuation characteristics. 5.The AI-based display screen layout control method of claim 4, wherein, The long-term trend characteristics and the short-term fluctuation characteristics are comprehensively evaluated to obtain attention weight values of each layout element in the layout element set, and the visual saliency ranking of each layout element is determined, including: The long-term trend characteristics and the short-term fluctuation characteristics are hierarchically processed using a pre-established scene classification standard to form a hierarchical behavior frequency distribution data set; The behavior frequency distribution data set is matched with a pre-established behavior frequency-weight mapping table to obtain the attention weight values of each layout element for the user; The long-term trend characteristics and the short-term fluctuation characteristics are compared in multiple layers to determine the repeated behavior frequency distribution characteristics across scenes; The attention weight values and the repeated behavior frequency distribution characteristics are combined to comprehensively score each layout element, and the visual saliency ranking of each layout element is determined according to the score from high to low. 6.The AI-based display screen layout control method of claim 5, wherein, The attention weight values are compared with a preset weight threshold to obtain high-priority elements, and the priority values of the high-priority elements are calculated based on the visual saliency ranking to obtain a preliminary layout scheme, including: Layout elements with attention weight values exceeding a preset weight threshold are marked as high-priority elements; According to the visual saliency ranking, the interaction areas corresponding to the top two layout elements are extracted, and the gaze dwell time of users in the interaction areas is counted within a preset monitoring time window. Regions with gaze dwell time exceeding a preset dwell time threshold are determined as user focus distribution; The user focus distribution and the user operation habit data are fused to obtain a comprehensive feature vector, and the priority values of the high-priority elements are calculated based on the comprehensive feature vector; The priority value corresponds to a proportion adjustment parameter obtained from a pre-established adjustment parameter database; The display position and size proportion of the high-priority elements are optimized according to the proportion adjustment parameter to obtain the preliminary layout scheme. 7.The AI-based display screen layout control method of claim 6, wherein, The preliminary layout scheme is adjusted to obtain an adjusted layout scheme, and the size and position of the layout elements are adaptively transformed based on the adjusted layout scheme to obtain an optimized display screen layout, including: When the proportion adjustment parameter set in the preliminary layout scheme exceeds a preset parameter threshold, the size transformation and position adjustment amplitude of the high-priority elements are limited to obtain an adjusted layout scheme; Based on the adjusted layout scheme, the size and position of each layout element are checked again. When the checking result shows that part of the layout elements overlap, the overlapping region is smoothed to obtain final transformation parameters; According to the final transformation parameters, the actual presentation mode of the layout elements is generated; According to the actual presentation mode, the priority value is adjusted to adapt to the resolution and scale limit of the user display screen, and an optimized display screen layout is obtained.

8. An AI-based display screen layout control system, characterized by, The method comprises: An interaction data collection module acquires operation records and focus point distribution data of a user in different scenarios from a user interaction environment, and stores the data to obtain a time series data stream of user behavior; A behavior rule analysis module extracts operation frequency and focus point migration path of a user in different interaction scenarios according to the time series data stream, and obtains dynamic change rule characteristics of user behavior; A behavior feature layering module mines user interaction records based on the dynamic change rule characteristics, classifies and arranges the user interaction records according to time and scenario dimensions to obtain a layered interaction data set, and determines long-term trend characteristics and short-term fluctuation characteristics of a user behavior mode according to the layered interaction data set; An element weight evaluation module determines a layout element set corresponding to different scenarios under a pre-established scenario classification standard, comprehensively evaluates the long-term trend characteristics and the short-term fluctuation characteristics, obtains focus weight values of each layout element in the layout element set, and determines a visual saliency ranking of each layout element; A layout decision generation module compares the focus weight values with a preset weight threshold to obtain high-priority elements, calculates priority values of the high-priority elements in combination with the visual saliency ranking, and obtains a preliminary layout scheme; A layout optimization adjustment module performs amplitude adjustment according to the preliminary layout scheme to obtain an adjusted layout scheme, and performs adaptive transformation on the size and position of layout elements based on the adjusted layout scheme, and obtains an optimized display screen layout.

9. An electronic device, comprising: The computer readable storage medium comprises a stored computer program, wherein when the computer program runs, the computer readable storage medium controls the device where the computer readable storage medium is located to execute the AI-based display screen layout control method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer readable storage medium comprises a stored computer program, wherein when the computer program runs, the computer readable storage medium controls the device where the computer readable storage medium is located to execute the AI-based display screen layout control method according to any one of claims 1 to 7.

Images