Interface visual hierarchy optimization method based on eye tracking

Abstract

This invention relates to the field of human-computer interaction and interface optimization, and discloses a method for optimizing the visual hierarchy of an interface based on eye tracking. The method includes: acquiring a set of interface elements of the interface to be optimized, and constructing a target visual hierarchy map based on a preset interaction task flow, the functional type of each interface element, and its information priority; collecting eye tracking data during the user's execution of a preset interaction task, the eye tracking data including first fixation time, cumulative fixation duration, number of regressions, and saccade shift relationships; determining the perception priority parameters of each interface element based on the eye tracking data, and constructing an actual perception hierarchy map; comparing the actual perception hierarchy map with the target visual hierarchy map to identify hierarchy deviation element groups and their corresponding deviation types; and calling the corresponding visual expression parameter reconstruction rules according to the deviation type to locally reconstruct the hierarchy deviation element groups while maintaining the preset interface structure constraints.

Description

Technical Field

[0001] This invention relates to the field of human-computer interaction and interface optimization, specifically to a method for optimizing the visual hierarchy of an interface based on eye tracking. Background Technology

[0002] As software interfaces are increasingly used in scenarios such as human-computer interaction, industrial monitoring, business approval, and information management, these interfaces often integrate multiple types of content, including status information, operation controls, feedback prompts, and auxiliary instructions. To ensure that users can quickly identify key information and complete corresponding operations according to the predetermined task flow, it is usually necessary to rationally design the visual hierarchy of each element in the interface.

[0003] In existing technologies, interface visual optimization typically relies on design experience, usability test results, or user eye-tracking heatmap distribution to adjust the color, size, position, or grouping of interface elements. While such solutions can reflect the user's attention area to some extent, most only remain at the level of overall evaluation, local salience enhancement, or heatmap statistics. They lack a clear depiction of the correspondence between the expected level of interface design and the actual level of user perception, and they also lack targeted optimization mechanisms that maintain the stability of the main task flow and the main interface structure.

[0004] Therefore, the existing technology has the following problems: due to the lack of technical means to compare the target visual level with the actual perception level based on the preset interactive task flow and to carry out local reconstruction accordingly, the expected entry order of key information in the interface is easily inconsistent with the user's actual attention process, which in turn causes visual level imbalance, task path interference and insufficient targeted interface optimization. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides an interface visual hierarchy optimization method based on eye tracking, in order to solve the technical problems existing in the prior art.

[0006] The above-mentioned technical objective of the present invention is achieved through the following technical solution: The eye-tracking-based method for optimizing the visual hierarchy of an interface includes the following steps: *S1: Obtain the set of interface elements of the interface to be optimized, and construct the target visual hierarchy diagram corresponding to the set of interface elements based on the preset interactive task flow corresponding to the interface to be optimized, the functional type of each interface element and the information priority. The target visual hierarchy diagram is used to represent the expected attention entry order and hierarchical relationship of each interface element in the preset interactive task flow. S2: Show the user the interface to be optimized and collect eye-tracking data during the user’s execution of preset interactive tasks. The eye-tracking data includes the first fixation time, cumulative fixation duration, number of regressions, and saccade shift relationships for each interface element. S3: Based on the first gaze time, cumulative gaze duration, number of re-gazes, and saccade shift relationship of each interface element, determine the perception priority parameters corresponding to each interface element, and construct the actual perception hierarchy map corresponding to the interface to be optimized based on the perception priority parameters and saccade shift relationship of each interface element. The actual perception hierarchy map is used to characterize the actual attention entry order and hierarchical relationship formed by the user for each interface element. S4: Compare the actual perception hierarchy map with the target visual hierarchy map, identify the hierarchy deviation element group, and determine the deviation type corresponding to the hierarchy deviation element group; S5: Based on the hierarchical deviation element group and the corresponding deviation type, call the visual expression parameter reconstruction rule corresponding to the deviation type. Under the condition of maintaining the preset interface structure constraints, perform local reconstruction of the visual expression parameters corresponding to the hierarchical deviation element group to obtain an optimized interface. S6: Perform eye-tracking data collection on the optimized interface again, construct the optimized actual perception hierarchy map, and compare the optimized actual perception hierarchy map with the target visual hierarchy map; when the number of the optimized hierarchy deviation element group is less than the number of the hierarchy deviation element group before reconstruction, the optimized interface is determined as the target optimized interface; when the number of the optimized hierarchy deviation element group is not less than the number of the hierarchy deviation element group before reconstruction, return to step S5.

[0007] Preferably, step S1 includes: Based on the preset interactive task flow, determine the corresponding interface elements in the task initiation stage, task execution stage, and task confirmation stage. Determine the expected attention entry order of each interface element according to the order in which they appear in the preset interactive task flow. Determine the visual hierarchy of each interface element based on its functional type and information priority. Based on the expected order of attention and the visual hierarchy to which each interface element belongs, establish the hierarchical relationship between each interface element in the target visual hierarchy diagram.

[0008] Preferably, in step S3, determining the perception priority parameter corresponding to each interface element based on the first gaze time, cumulative gaze duration, number of re-gazes, and saccade transfer relationship of each interface element includes: The first entry order value of each interface element is determined based on the first gaze time of each interface element. The dwell intensity value of each interface element is determined based on the cumulative gaze duration of each interface element. Based on the number of times each interface element is viewed, determine the repetition attention value of each interface element; Based on the saccade transfer relationship, determine the transfer association value between each interface element and other interface elements; Based on the initial entry sequence value, dwell intensity value, repeated attention value, and transfer association value, the perception priority parameter corresponding to each interface element is determined.

[0009] Preferably, in step S3, constructing the actual perception hierarchy map corresponding to the interface to be optimized based on the perception priority parameters of each interface element and the saccade transfer relationship includes: The actual order in which attention enters each interface element is determined according to the perceived priority parameter of each interface element. Based on the described scanning transfer relationship, determine the actual transfer connection relationship between each interface element; Based on the actual attention entry order and actual transfer connection relationship of each interface element, construct an actual perception hierarchy diagram.

[0010] Preferably, in step S4, the types of deviations include first gaze order deviation, dwell competition deviation, retrospective interference deviation, cross-layer jump deviation, and same-layer misleading deviation; The first gaze order deviation is used to characterize the inconsistency between the interface element that is actually gazed first in the same task phase and the interface element that is expected to enter the attention process first in the target visual hierarchy diagram. The dwell competition bias is used to characterize the cumulative gaze duration of low-visual-level interface elements that is greater than the cumulative gaze duration of high-visual-level interface elements that have a hierarchical relationship with them. The back-look interference deviation is used to characterize the number of times a user looks back at the corresponding interface element of the completed task stage after completing the current task stage. The cross-layer jump deviation is used to characterize the situation in which the user's saccade shift relationship skips the intermediate visual layer and directly jumps to the interface element of a non-adjacent visual layer. The same-layer misleading bias is used to characterize the fact that the perception priority parameter of non-target interface elements is higher than that of target interface elements within the same visual layer.

[0011] Preferably, in step S5, the preset interface structure constraints include the following constraints: The boundary assignment relationships of each functional zone remain unchanged; The sequential relationship between the interface elements in the main operation chain remains unchanged; The number of primary functional partitions remains unchanged; The correspondence between core input control elements and their corresponding feedback prompt elements remains unchanged.

[0012] Preferably, in step S5, the visual expression parameters include element display size, element spacing, contrast parameters, partition boundary display parameters, and text information density; The partition boundary display parameters are used to characterize the display width of the boundary lines between the functional partitions of the interface and the difference in brightness between the boundary lines and the background.

[0013] Preferably, in step S5, the invocation of the visual expression parameter reconstruction rule corresponding to the deviation type includes: When the deviation type is first-look order deviation, the element display size and contrast parameters of the target interface element in the hierarchical deviation element group are adjusted. When the deviation type is a dwell competition deviation, the text information density and element spacing of non-target interface elements in the hierarchical deviation element group are adjusted. When the deviation type is backview interference deviation, adjust the partition boundary display parameters of the corresponding interface elements in the completed task stage; When the deviation type is cross-layer jump deviation, the display parameters of the partition boundary and the element spacing between adjacent visual layers are adjusted; When the deviation type is same-layer misleading deviation, the contrast parameters of target interface elements and non-target interface elements within the same visual layer are adjusted differently.

[0014] Preferably, in step S5, the local reconstruction is limited to adjusting only the visual expression parameters corresponding to the hierarchical deviation element group, without changing the display position and function type of the interface elements that are not identified as hierarchical deviation element groups.

[0015] Preferably, in step S6, based on the fact that the number of hierarchical deviation element groups after optimization is less than the number of hierarchical deviation element groups before reconstruction, when the actual attention entry order of the target interface elements in the optimized actual perception hierarchical diagram is consistent with the expected attention entry order in the target visual hierarchical diagram, the optimized interface is determined as the target optimized interface.

[0016] In summary, the present invention has the following main beneficial effects: This application constructs a target visual hierarchy map based on a preset interactive task flow, interface element function types, and information priority. It further combines this with eye-tracking data from the user's eye movements during the preset interactive task to construct an actual perception hierarchy map, achieving a correspondence between the interface design expectations and the user's actual perception process. By comparing the target visual hierarchy map with the actual perception hierarchy map, previously difficult-to-identify attention entry sequence deviations, hierarchy association deviations, and visual competition relationships in the interface can be specifically located to the hierarchy deviation element groups. This avoids making general adjustments to the interface based solely on overall scoring results, local hotspot distribution, or subjective design experience, thus improving the targeting and feasibility of interface visual hierarchy optimization.

[0017] By further differentiating the hierarchical deviation element groups into first-look order deviation, dwell competition deviation, back-look interference deviation, cross-layer jump deviation, and same-layer misleading deviation, and while maintaining the preset interface structure constraints, the system invokes visual expression parameter reconstruction rules corresponding to different deviation types. This achieves the effect of implementing local optimization without disrupting the main interface structure and task flow. By adjusting only the element display size, element spacing, contrast parameters, partition boundary display parameters, and text information density corresponding to the hierarchical deviation element groups, the interface optimization process always revolves around the key interface elements causing hierarchical imbalance and their relationships. This avoids the task path confusion and structural instability problems caused by overall redesign, arbitrary changes to partition relationships, or alterations to the core operation order in existing technologies.

[0018] By collecting eye-tracking data again after local reconstruction, and comparing the optimized actual perceptual hierarchy map with the target visual hierarchy map, the number of hierarchical deviation element groups and the actual attention entry order of target interface elements are verified, achieving a closed-loop confirmation of the interface optimization results. This reconstruction-re-verification process ensures that interface visual hierarchy optimization is no longer a one-time adjustment, but rather a continuous process of target establishment, deviation identification, local reconstruction, and verification. This improves the consistency between the optimization results and the preset interactive task flow, enhances the priority identification ability of key information in the interface, and strengthens the hierarchical clarity of the task execution process. Attached Figure Description

[0019] Figure 1 This is a flowchart of the method of the present invention. Detailed Implementation

[0020] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0021] Example 1 refer to Figure 1 An eye-tracking-based method for optimizing the visual hierarchy of an interface includes the following steps: S1: Obtain the set of interface elements of the interface to be optimized, and construct the target visual hierarchy diagram corresponding to the set of interface elements based on the preset interactive task flow corresponding to the interface to be optimized, the functional type of each interface element and the information priority. The target visual hierarchy diagram is used to represent the expected attention entry order and hierarchical relationship of each interface element in the preset interactive task flow. S2: Show the user the interface to be optimized and collect eye-tracking data during the user’s execution of preset interactive tasks. The eye-tracking data includes the first fixation time, cumulative fixation duration, number of regressions, and saccade shift relationships for each interface element. S3: Based on the first gaze time, cumulative gaze duration, number of re-gazes, and saccade shift relationship of each interface element, determine the perception priority parameters corresponding to each interface element, and construct the actual perception hierarchy map corresponding to the interface to be optimized based on the perception priority parameters and saccade shift relationship of each interface element. The actual perception hierarchy map is used to characterize the actual attention entry order and hierarchical relationship formed by the user for each interface element. S4: Compare the actual perception hierarchy map with the target visual hierarchy map, identify the hierarchy deviation element group, and determine the deviation type corresponding to the hierarchy deviation element group; S5: Based on the hierarchical deviation element group and the corresponding deviation type, call the visual expression parameter reconstruction rule corresponding to the deviation type. Under the condition of maintaining the preset interface structure constraints, perform local reconstruction of the visual expression parameters corresponding to the hierarchical deviation element group to obtain an optimized interface. S6: Perform eye-tracking data collection on the optimized interface again, construct the optimized actual perception hierarchy map, and compare the optimized actual perception hierarchy map with the target visual hierarchy map; when the number of the optimized hierarchy deviation element group is less than the number of the hierarchy deviation element group before reconstruction, the optimized interface is determined as the target optimized interface; when the number of the optimized hierarchy deviation element group is not less than the number of the hierarchy deviation element group before reconstruction, return to step S5.

[0022] S1. Obtain the set of UI elements of the interface to be optimized and construct the target visual hierarchy diagram: In this embodiment, the first step is to obtain a set of interface elements for the interface to be optimized. The interface elements in this set refer to visual objects within the interface that can be visually perceived by the user and participate in a preset interactive task flow. These interface elements may include text elements, icon elements, input control elements, selection control elements, feedback prompt elements, operation button elements, and functional area boundary elements. These interface elements can be obtained by parsing an interface configuration file, a front-end rendering tree, a layout description file, or a runtime interface object tree. As long as the identifier, display area, category, and task association of the interface elements can be determined, it can serve as the basis for subsequent processing.

[0023] After obtaining the set of interface elements, a target visual hierarchy diagram corresponding to the set of interface elements is constructed based on the preset interactive task flow corresponding to the interface to be optimized, the functional type of each interface element, and the information priority.

[0024] The preset interactive task flow is predefined by a task configuration file. The task configuration file records interface element identifiers, task stage identifiers, processing order within stages, and transition order between stages. When the same interface corresponds to multiple interactive tasks, a corresponding task configuration file is created for each interactive task, and a target visual hierarchy graph is constructed based on the task configuration file corresponding to the current task to be optimized. For multiple interface elements within the same visual hierarchy, the task configuration file also records the target interface elements within that visual hierarchy that are expected to preferentially enter the attention process.

[0025] The preset interactive task flow is not an abstract task description, but a specific operation flow corresponding to the interface to be optimized. For example, in the alarm handling interface, the preset interactive task flow may include: reading the current alarm level, viewing the alarm source, confirming the alarm details, entering the handling instructions, executing the handling command, and viewing the handling feedback. In other words, the preset interactive task flow is used to limit the user's focus order on the interface when normally completing business operations, rather than merely expressing the interface design intent.

[0026] The function type is used to characterize the role of interface elements in the interaction process. The function type can include information reading, input operation, confirmation feedback, navigation switching, and auxiliary instructions. The function type can be directly determined by the component category in the interface design file, front-end component identifiers, back-end configuration tags, or task configuration files.

[0027] The information priority is pre-configured using a discrete level approach. In this embodiment, information priority is divided into first-level, second-level, and third-level information priority. First-level information priority corresponds to interface elements that directly affect the start or end of the current task phase; second-level information priority corresponds to interface elements that affect the specific processing of the current task phase; and third-level information priority corresponds to auxiliary or historical viewing interface elements. The processing unit reads the information priority level corresponding to each interface element from the task configuration file and constructs a target visual hierarchy diagram based on the preset interactive task flow. Using a discrete level configuration method ensures that information priority has a clear source and defined boundaries, avoiding reliance solely on subjective design experience.

[0028] In the specific construction process, the processing unit first divides the preset interactive task flow into three stages: task initiation, task execution, and task confirmation, based on the task configuration file. Then, it maps the interface elements corresponding to the task initiation, task execution, and task confirmation stages to separate sets of interface elements. Next, based on the order in which each interface element appears in the preset interactive task flow, the expected attention entry order of each interface element is determined; finally, considering the functional type and information priority of each interface element, the visual hierarchy to which each interface element belongs is determined.

[0029] In this embodiment, the target visual hierarchy graph is represented by a directed graph structure. Nodes in the graph correspond to interface elements, and directed edges represent the expected attention transfer relationship from one interface element to the next in the preset interactive task flow. If two interface elements need to be paid attention to sequentially in the same task stage, their order relationship is recorded in the target visual hierarchy graph; if two interface elements belong to the same level in the same task stage, their co-level relationship is recorded in the target visual hierarchy graph. The task configuration file also specifies the target interface elements within the same level. Therefore, the target visual hierarchy graph records both the expected attention entry order of each interface element and the hierarchical relationship between each interface element.

[0030] In this way, the target visual hierarchy map is not derived from eye-tracking data, but is a priori based on task configuration files, function types, and information priorities. This is fundamentally different from methods that merely score the overall design or directly determine important areas based on heat map distribution.

[0031] S2. Display the interface to be optimized and collect eye-tracking data: In this embodiment, the interface to be optimized is displayed to the user, and eye-tracking data is collected during the user's execution of a preset interactive task. The eye-tracking data includes the first fixation time, cumulative fixation duration, number of regressions, and saccade shift relationships for each interface element.

[0032] Among them, first fixation time refers to the time when a user first fixates on the display area corresponding to a certain interface element; cumulative fixation duration refers to the sum of all fixation times of the user on the display area corresponding to the interface element during task execution; number of back-looks refers to the number of times the user returns to fixate on the interface element after completing the first fixation on the interface element and moving to other interface elements; saccade shift relationship refers to the directional relationship of the user's gaze shifting from the display area corresponding to one interface element to the display area corresponding to another interface element.

[0033] In its implementation, the eye-tracking acquisition unit first calibrates the user's gaze to establish a correspondence between the screen coordinate system and the eye-tracking sampling coordinate system. After calibration, when the user performs a preset interactive task, the unit continuously samples and records the sequence of the user's gaze points on the interface. The processing unit matches the gaze point sequence with the display area occupied by each interface element to obtain the first fixation time, cumulative fixation duration, number of regressions, and saccade transfer relationships for each interface element.

[0034] It should be noted that this application does not use individual eye-tracking sampling points as the basis for optimization. Instead, it merges discrete eye-tracking sampling points to the interface element level to form interface element-level eye-tracking features. This allows subsequent calculations of perception priority parameters, hierarchical deviation identification, and visual expression parameter reconstruction to directly apply to interface elements, rather than to scattered eye-tracking coordinate points, thereby facilitating the coordination between task flow, eye-tracking behavior, and interface reconstruction.

[0035] S3. Determine the perception priority parameters and construct the actual perception hierarchy diagram: In this embodiment, based on the first gaze time, cumulative gaze duration, number of re-gazes, and saccade transfer relationships of each interface element, the perception priority parameters corresponding to each interface element are determined, and based on the perception priority parameters and saccade transfer relationships of each interface element, an actual perception hierarchy map corresponding to the interface to be optimized is constructed.

[0036] Let the set of UI elements to be optimized be:

[0037] in, Indicates the first Interface elements, This indicates the total number of UI elements. (For UI elements...) Define its first entry order value Residence intensity value Repeated attention value and transfer associated values .

[0038] First entry sequence value: Let Represents UI elements The first entry sequence number among all interface elements; the smaller the sequence number, the earlier it enters the user's attention process. Therefore, the first entry sequence value... Represented as:

[0039] in, The larger the value, the more important it is for the UI element. The earlier you get into the user's attention process.

[0040] Residence intensity value: Let Represents UI elements The cumulative gaze duration, then the dwell intensity value Represented as:

[0041] in, Represents UI elements Percentage of time spent on all UI elements.

[0042] Repeated attention value: set Represents UI elements The number of times the view is repeated, then the repeat attention value is... Represented as:

[0043] Adding 1 to the denominator is to prevent division by zero when all interface elements have a view count of 0. The larger the value, the more obvious the user's repeated attention to the interface element.

[0044] Transfer associated values: Let Indicates that the user's gaze is directed through the interface elements. Transfer to UI elements The number of times, then the interface element Transfer correlation value Represented as:

[0045] Among them, it is used to characterize the degree of concentration of gaze shift from other interface elements to the current interface element, so as to reflect the position of the current interface element in the visual path of the interface.

[0046] Perception priority parameters: In this embodiment, interface elements Perception priority parameters Represented as:

[0047] in, , , and These are preset weighting coefficients, all of which are constants greater than 0; Used to characterize the weight of the first entry order value on the perception priority parameter. The weight used to characterize the influence of dwell intensity value on the perception priority parameter. The weight used to characterize the influence of transfer association values ​​on the perception priority parameter. Used to characterize the suppression weight of repeated attention values ​​on the perceived priority parameter.

[0048] In this embodiment, the weighting coefficient , , and The parameters are pre-set in the configuration file before interface optimization begins and remain unchanged throughout the same optimization process. The configuration file is differentiated based on the interface type. For interfaces primarily focused on quickly identifying key information, the parameters are optimized to improve… The configuration weight; for interfaces primarily used for continuous reading and verification, increase... The configuration weight; for interfaces that primarily handle workflow transitions, increase... The configuration weight; for interfaces with significant back-view interference, increase The configuration weights. Those skilled in the art can select the corresponding parameter configuration group according to the interface type, and keep the weight coefficients unchanged in the same optimization process to ensure that the calculation method of the perception priority parameter is consistent.

[0049] The reason for subtracting the repetition rate from the perception priority parameter is that a high number of re-views usually indicates that the user is re-confirming, repeatedly recognizing, or looking back after being distracted by the interface element. If this is not differentiated from the first fixation time, cumulative fixation duration, and the correlation with the transition, it is easy to misjudge high-priority objects caused by distraction. This application uses the above calculation method to map interface elements that enter the attention process early, remain stable, have clear connections, and are not repeatedly re-viewed as high-perception priority parameter objects.

[0050] After obtaining the perception priority parameters corresponding to each interface element, the processing unit determines the actual attention entry order of each interface element according to the size of the perception priority parameters; then, it determines the actual transfer connection relationship between interface elements according to the saccade transfer relationship; finally, it constructs the actual perception hierarchy map based on the actual attention entry order and the actual transfer connection relationship.

[0051] In this embodiment, the actual perception hierarchy map and the target visual hierarchy map are represented by the same graph structure, both using a directed graph structure, which facilitates subsequent deviation identification on a node-by-node, edge-by-edge, and level-by-level basis. This is significantly different from processing methods that only output the overall score or only output the heat map without establishing a perception hierarchy map corresponding to the task flow.

[0052] S4. Identify the hierarchical deviation element group and determine the deviation type: In this embodiment, the actual perceived hierarchy map is compared with the target visual hierarchy map to identify the hierarchy deviation element group and determine the deviation type corresponding to the hierarchy deviation element group.

[0053] The aforementioned hierarchical deviation element group refers to a group of one or more related interface elements identified during the same comparison process that exhibit hierarchical inconsistencies, sequential inconsistencies, significant backtracking interference, or abnormal cross-layer transitions compared to the target visual hierarchy map. The reason for using element groups as the identification object is that many interface hierarchy problems are not caused by isolated single interface elements, but rather by the combined effects of target interface elements, interfering interface elements, adjacent hierarchy interface elements, and completed stage interface elements. Identifying them as element groups facilitates subsequent targeted reconstruction.

[0054] In this embodiment, the types of deviations include first gaze order deviation, dwell competition deviation, retrospective interference deviation, cross-layer jump deviation, and same-layer misleading deviation.

[0055] Among them, the first fixation order bias is used to characterize the inconsistency between the interface element that is actually fixated first in the same task phase and the interface element that is expected to enter the attention process first in the target visual hierarchy diagram.

[0056] Dwell competition bias is used to characterize the cumulative gaze duration of lower visual hierarchy interface elements that is greater than the cumulative gaze duration of higher visual hierarchy interface elements that have a hierarchical relationship with them.

[0057] The back-look interference bias is used to characterize the number of times a user looks back at the corresponding interface element of the completed task stage after completing the current task stage, which exceeds a preset threshold.

[0058] Cross-level jump deviation is used to characterize situations in which a user's saccade transitions skip intermediate visual levels and directly jumps to interface elements in non-adjacent visual levels.

[0059] The same-layer misleading bias is used to characterize the perception priority parameter of non-target interface elements being higher than that of target interface elements within the same visual layer.

[0060] The target interface element within the same visual level refers to the interface element configured in the target visual level diagram as expected to be the first to enter the attention process within that visual level. This configuration relationship is pre-recorded by the task configuration file.

[0061] The preset threshold number of attempts is determined before the current interface to be optimized begins optimization and remains unchanged during the same round of optimization for the current interface. When historical interaction logs exist, the preset threshold number of attempts is determined based on historical interaction logs of similar interfaces; when no historical interaction logs exist, the preset threshold number of attempts is determined based on the test results of sample users on the baseline interface; when neither historical interaction logs nor baseline test results exist, the preset threshold number of attempts in the system's default configuration is read. Through the above determination method, the threshold number in the backview interference deviation can have a clear source and fixed boundaries.

[0062] When identifying hierarchical deviation element groups, the processing unit first compares the expected attention entry order with the actual attention entry order of each node in the target visual hierarchy map and the actual perception hierarchy map one by one. Then, it compares the hierarchical association relationship with the actual transfer connection relationship. Finally, it combines the cumulative fixation duration and the number of retrospectives to form a set of deviation elements. Subsequently, according to the association relationship in the map, the mutually influential deviation elements are merged into the corresponding hierarchical deviation element groups, and the deviation type is marked for each hierarchical deviation element group.

[0063] Through the above method, this application does not only obtain an overall score result, but also locates the deviation to a specific hierarchical deviation element group and its deviation type, providing a basis for subsequent constrained local reconstruction.

[0064] S5. Perform local reconstruction under structural constraints: In this embodiment, based on the hierarchical deviation element group and the corresponding deviation type, the visual expression parameter reconstruction rule corresponding to the deviation type is invoked. Under the condition of maintaining the preset interface structure constraints, the visual expression parameters corresponding to the hierarchical deviation element group are locally reconstructed to obtain an optimized interface.

[0065] The preset interface structure constraints include: the boundary attribution relationship of each functional area remains unchanged; the sequential position relationship of each interface element in the main operation chain remains unchanged; the number of first-level functional areas remains unchanged; and the correspondence between core input control elements and corresponding feedback prompt elements remains unchanged.

[0066] Specifically, the boundary relationships of each functional area remain unchanged, meaning that a certain interface element belongs to the same functional area before optimization and remains so after optimization, and migration across functional areas is not allowed; the sequential relationship of each interface element in the main operation chain remains unchanged, meaning that the order of the core interface elements constituting the critical task flow is not reversed; the number of first-level functional areas remains unchanged, meaning that the optimization process does not change the main structure of the interface by adding or deleting main areas; and the correspondence between core input control elements and their corresponding feedback prompts remains unchanged, meaning that the error prompts, confirmation prompts, or status prompts corresponding to a certain input control are still bound to that input control.

[0067] In this embodiment, the visual expression parameters include element display size, element spacing, contrast parameters, partition boundary display parameters, and text information density.

[0068] Among them, element display size is used to characterize the display area of ​​interface elements in the interface; element spacing is used to characterize the display distance between interface elements and adjacent interface elements; contrast parameter is used to characterize the degree of brightness contrast between interface elements and the background or surrounding interface elements; partition boundary display parameter is used to characterize the display width of the boundary line between interface functional partitions and the brightness difference between the boundary line and the background; text information density is used to characterize the amount of text content and the compactness of text arrangement within a unit display area.

[0069] The partition boundary display parameters are directly read from the boundary line width and boundary brightness difference fields in the interface style configuration file, and then written back to the corresponding configuration fields by the processing unit to complete the adjustment. The text information density is adjusted by modifying the text block width, line spacing, paragraph spacing, and auxiliary description folding status. In the same local reconstruction, the text information content itself remains unchanged.

[0070] In this embodiment, there is a correspondence between the deviation type and the visual expression parameter reconstruction rule, as detailed below: When the deviation type is first-look order deviation, the display size and contrast parameters of the target interface elements in the hierarchical deviation element group are adjusted so that the target interface elements enter the user's attention process earlier while maintaining the original task position relationship.

[0071] When the deviation type is dwell competition deviation, the text information density and element spacing of non-target interface elements in the hierarchical deviation element group are adjusted to reduce the unreasonable dwell competition formed by low-level interface elements on high-level interface elements.

[0072] When the deviation type is backview interference deviation, the display parameters of the partition boundary of the interface element corresponding to the completed task stage are adjusted to reduce the visual interference of the completed stage area on the current stage processing.

[0073] When the deviation type is cross-layer jump deviation, the display parameters of the partition boundary and the element spacing between adjacent visual layers are adjusted to enhance the visual connection between adjacent layers.

[0074] When the deviation type is same-layer misleading deviation, the contrast parameters of target interface elements and non-target interface elements within the same visual layer are adjusted differently so that the target interface elements in the same layer are given priority in the user's attention process.

[0075] It should be noted that the above reconstruction is a partial reconstruction, meaning that only the visual expression parameters corresponding to the element groups identified as hierarchical deviations are adjusted, without changing the display position and function type of interface elements not identified as hierarchical deviation elements. In this way, this application eliminates hierarchical deviations without disrupting the main interface structure and main task flow.

[0076] S6. After reconstruction, verify again and determine the target optimized interface: In this embodiment, eye-tracking data is collected again on the optimized interface to construct an optimized actual perception hierarchy map, and the optimized actual perception hierarchy map is compared with the target visual hierarchy map.

[0077] When the number of hierarchical deviation element groups after optimization is less than the number of hierarchical deviation element groups before reconstruction, the optimized interface is selected as a candidate optimized interface. Based on this, it is further determined whether the actual attention entry order of the target interface elements in the optimized actual perceptual hierarchy map is consistent with the expected attention entry order in the target visual hierarchy map. When they are consistent, the optimized interface is determined as the target optimized interface.

[0078] When the number of optimized hierarchical deviation element groups is not less than the number of hierarchical deviation element groups before reconstruction, or when the number of hierarchical deviation element groups is reduced but the actual attention entry order of the target interface elements is still not consistent with the expected attention entry order, then return to step S5 and continue to perform local reconstruction on the visual expression parameters corresponding to the hierarchical deviation element groups.

[0079] In this embodiment, the judgment criterion for S6 is not whether the overall interface score has improved, nor whether the overall click efficiency has improved, but rather whether the hierarchical deviation has converged and whether the actual attention entry order of the target interface elements has returned to the expected order. Thus, this application forms a closed-loop implementation logic encompassing target visual hierarchy map establishment, actual perception hierarchy map construction, hierarchical deviation element group identification, local reconstruction, and post-reconstruction verification.

[0080] Example 2 This embodiment further explains the key terms in Embodiment 1 to avoid ambiguity and enhance implementation support.

[0081] Interface elements refer to display objects that are individually identifiable within the interface layout structure, visually perceptible to the user, and capable of participating in predefined interactive task flows. The boundaries of an interface element are directly determined by its coordinate range in the layout file, the bounding box range in the rendering tree, or the runtime drawing area. If an interface object consists of multiple tightly coupled sub-elements and is typically perceived and processed as a whole during task execution, it is treated as a single interface element; if its sub-elements perform different functions in the task flow, they are treated as separate interface elements.

[0082] The task configuration file is a data file pre-created before the interface optimization begins, used to record the task flow definition corresponding to the interface to be optimized. The task configuration file must at least record the interface element identifier, task stage identifier, intra-stage order, inter-stage order, and identifiers of target interface elements at the same level. The processing unit directly reads the task configuration file in step S1 to avoid inconsistencies in the target visual hierarchy diagram construction standards caused by real-time manual judgment.

[0083] The parameter configuration file is a data file created before the interface optimization begins. It records the weight coefficient configuration group corresponding to the current interface type. The parameter configuration file records the weight coefficients. The processing unit directly reads the weight configuration group corresponding to the current interface type to be optimized in step S3, and keeps this configuration group unchanged in the same optimization process.

[0084] The preset threshold for the number of iterations is determined before the current interface to be optimized begins optimization. The priority method for determination is as follows: first, it is determined based on historical interaction logs of similar interfaces; if no historical interaction logs are available, it is determined based on the test results of sample users on the baseline interface; if neither exists, the system default configuration is read. The processing unit consistently uses the same preset threshold for the number of iterations during the same round of optimization for the current interface.

[0085] The partition boundary display parameters specifically include the display width of the boundary line and the brightness difference between the boundary line and the background. The processing unit adjusts these parameters by modifying the corresponding fields in the interface style configuration file. Therefore, the partition boundary display parameters have a clearly defined implementation target and a direct modification path.

[0086] Text information density refers to the amount of text content and the compactness of text arrangement within a unit display area. The processing unit adjusts text information density by modifying text block width, line spacing, paragraph spacing, and the collapsible state of auxiliary descriptions. Within the same local reconstruction, the text content itself remains unchanged to avoid confusing content changes with visual hierarchy optimization.

[0087] Example 3 The following describes the specific application process of this application in conjunction with the alarm handling interface.

[0088] Assume the interface to be optimized includes a current alarm level display area, an alarm source display area, a handling description input area, a handling button area, a history view area, an auxiliary help area, and a handling feedback area. For this interface, the preset interactive task flow recorded in the task configuration file is: read the current alarm level, read the alarm source, input handling description, execute the handling button, and view the handling feedback. Specifically, the current alarm level display area and the alarm source display area correspond to the task start stage, the handling description input area and the handling button area correspond to the task execution stage, the handling feedback area corresponds to the task confirmation stage, and the history view area and the auxiliary help area correspond to auxiliary description interface elements.

[0089] In step S1, the processing unit constructs a target visual hierarchy diagram based on the task configuration file, function type, and information priority, and sets the current alarm level display area as the target interface element within the task start phase, and sets the disposal button area as the target interface element within the task execution phase.

[0090] In step S2, the eye-tracking acquisition unit collects eye-tracking data during the user's execution of the alarm handling task, and the processing unit obtains the first fixation time, cumulative fixation duration, number of back-looks, and saccade transfer relationship for each interface element.

[0091] In step S3, the processing unit calculates the first entry order value, dwell intensity value, repeated attention value and transfer association value of each interface element based on the above data, and then calculates the perception priority parameters of each interface element according to the formula, and constructs the actual perception hierarchy map.

[0092] If, in step S4, it is found that the first gaze time in the history viewing area is earlier than that in the current alarm level display area, and its cumulative gaze duration is greater than that in the alarm source display area, and the user frequently looks back at the history viewing area after entering the handling instructions, then it indicates that there are simultaneous first gaze order deviation, dwell competition deviation, and lookback interference deviation in the interface. In this case, the current alarm level display area, the history viewing area, and its associated functional areas can be identified as a hierarchical deviation element group.

[0093] In step S5, the processing unit adjusts the element display size and contrast parameters of the current alarm level display area, adjusts the text information density and element spacing of the history viewing area, and adjusts the partition boundary display parameters of the history partition. During this process, the boundary relationships of each functional partition, the sequential positional relationships of interface elements in the main operation chain, the number of first-level functional partitions, and the correspondence between core input control elements and corresponding feedback prompt elements remain unchanged.

[0094] In step S6, eye-tracking data of the optimized interface is collected again, and an optimized actual perception hierarchy map is constructed. If the actual attention entry order of the current alarm level display area in the optimized actual perception hierarchy map is consistent with the expected attention entry order in the target visual hierarchy map, and the number of corresponding hierarchy deviation element groups is less than the number of hierarchy deviation element groups before reconstruction, then the optimized interface is determined to be the target optimized interface; otherwise, return to step S5 to continue local reconstruction.

[0095] As can be seen from the above implementation process, this application does not simply identify areas that are viewed more frequently as more important areas, nor does it directly enlarge or move interface elements based on overall heatmap results. Instead, it first establishes a target visual hierarchy map, then constructs an actual perceptual hierarchy map, subsequently identifies hierarchy deviation element groups, performs local reconstruction of these groups under structural constraints, and finally verifies the results through eye tracking again. Therefore, this application differs fundamentally from existing solutions that only perform eye-tracking evaluation, only perform attention heatmap analysis, or only perform overall layout comparison and adjustment, in terms of the target object, intermediate result processing, and optimized execution logic.

[0096] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for optimizing the visual hierarchy of an interface based on eye tracking, characterized in that, Includes the following steps: S1: Obtain the set of interface elements of the interface to be optimized, and construct the target visual hierarchy diagram corresponding to the set of interface elements based on the preset interactive task flow corresponding to the interface to be optimized, the functional type of each interface element and the information priority. The target visual hierarchy diagram is used to represent the expected attention entry order and hierarchical relationship of each interface element in the preset interactive task flow. S2: Show the user the interface to be optimized and collect eye-tracking data during the user’s execution of preset interactive tasks. The eye-tracking data includes the first fixation time, cumulative fixation duration, number of regressions, and saccade shift relationships for each interface element. S3: Based on the first gaze time, cumulative gaze duration, number of re-gazes, and saccade shift relationship of each interface element, determine the perception priority parameters corresponding to each interface element, and construct the actual perception hierarchy map corresponding to the interface to be optimized based on the perception priority parameters and saccade shift relationship of each interface element. The actual perception hierarchy map is used to characterize the actual attention entry order and hierarchical relationship formed by the user for each interface element. S4: Compare the actual perception hierarchy map with the target visual hierarchy map, identify the hierarchy deviation element group, and determine the deviation type corresponding to the hierarchy deviation element group; S5: Based on the hierarchical deviation element group and the corresponding deviation type, call the visual expression parameter reconstruction rule corresponding to the deviation type. Under the condition of maintaining the preset interface structure constraints, perform local reconstruction of the visual expression parameters corresponding to the hierarchical deviation element group to obtain an optimized interface. S6: Perform eye-tracking data collection on the optimized interface again, construct the optimized actual perception hierarchy map, and compare the optimized actual perception hierarchy map with the target visual hierarchy map; when the number of the optimized hierarchy deviation element group is less than the number of the hierarchy deviation element group before reconstruction, the optimized interface is determined as the target optimized interface; when the number of the optimized hierarchy deviation element group is not less than the number of the hierarchy deviation element group before reconstruction, return to step S5.

2. The interface visual hierarchy optimization method based on eye tracking according to claim 1, characterized in that, Step S1 includes: Based on the preset interactive task flow, determine the corresponding interface elements in the task initiation stage, task execution stage, and task confirmation stage. Determine the expected attention entry order of each interface element according to the order in which they appear in the preset interactive task flow. Determine the visual hierarchy of each interface element based on its functional type and information priority. Based on the expected order of attention and the visual hierarchy to which each interface element belongs, establish the hierarchical relationship between each interface element in the target visual hierarchy diagram.

3. The interface visual hierarchy optimization method based on eye tracking according to claim 2, characterized in that, In step S3, determining the perception priority parameters for each interface element based on its initial gaze time, cumulative gaze duration, number of regressions, and saccade shift relationships includes: The first entry order value of each interface element is determined based on the first gaze time of each interface element. The dwell intensity value of each interface element is determined based on the cumulative gaze duration of each interface element. Based on the number of times each interface element is viewed, determine the repetition attention value of each interface element; Based on the saccade transfer relationship, determine the transfer association value between each interface element and other interface elements; Based on the initial entry sequence value, dwell intensity value, repeated attention value, and transfer association value, the perception priority parameter corresponding to each interface element is determined.

4. The interface visual hierarchy optimization method based on eye tracking according to claim 3, characterized in that, In step S3, the construction of the actual perception hierarchy map corresponding to the interface to be optimized based on the perception priority parameters of each interface element and the saccade transfer relationship includes: The actual order in which attention enters each interface element is determined according to the perceived priority parameter of each element. Based on the described scanning transfer relationship, determine the actual transfer connection relationship between each interface element; Based on the actual attention entry order and actual transfer connection relationship of each interface element, construct an actual perception hierarchy diagram.

5. The interface visual hierarchy optimization method based on eye tracking according to claim 4, characterized in that, In step S4, the types of deviations include first gaze order deviation, dwell competition deviation, retrospective interference deviation, cross-layer jump deviation, and same-layer misleading deviation; The first gaze order deviation is used to characterize the inconsistency between the interface element that is actually gazed first in the same task phase and the interface element that is expected to enter the attention process first in the target visual hierarchy diagram. The dwell competition bias is used to characterize the cumulative gaze duration of low-visual-level interface elements that is greater than the cumulative gaze duration of high-visual-level interface elements that have a hierarchical relationship with them. The back-look interference deviation is used to characterize the number of times a user looks back at the corresponding interface element of the completed task stage after completing the current task stage. The cross-layer jump deviation is used to characterize the situation in which the user's saccade shift relationship skips the intermediate visual layer and directly jumps to the interface element of a non-adjacent visual layer. The same-layer misleading bias is used to characterize the fact that the perception priority parameter of non-target interface elements is higher than that of target interface elements within the same visual layer.

6. The interface visual hierarchy optimization method based on eye tracking according to claim 5, characterized in that, In step S5, the preset interface structure constraints include the following constraints: The boundary assignment relationships of each functional zone remain unchanged; The sequential relationship between the interface elements in the main operation chain remains unchanged; The number of primary functional zones remains unchanged; The correspondence between core input control elements and their corresponding feedback prompt elements remains unchanged.

7. The interface visual hierarchy optimization method based on eye tracking according to claim 6, characterized in that, In step S5, the visual expression parameters include element display size, element spacing, contrast parameters, partition boundary display parameters, and text information density; The partition boundary display parameters are used to characterize the display width of the boundary lines between the functional partitions of the interface and the difference in brightness between the boundary lines and the background.

8. The interface visual hierarchy optimization method based on eye tracking according to claim 7, characterized in that, In step S5, the invocation of the visual expression parameter reconstruction rule corresponding to the deviation type includes: When the deviation type is first-look order deviation, the element display size and contrast parameters of the target interface element in the hierarchical deviation element group are adjusted. When the deviation type is a dwell competition deviation, the text information density and element spacing of non-target interface elements in the hierarchical deviation element group are adjusted. When the deviation type is backview interference deviation, adjust the partition boundary display parameters of the corresponding interface elements in the completed task stage; When the deviation type is cross-layer jump deviation, the display parameters of the partition boundary and the element spacing between adjacent visual layers are adjusted; When the deviation type is same-layer misleading deviation, the contrast parameters of target interface elements and non-target interface elements within the same visual layer are adjusted differently.

9. The interface visual hierarchy optimization method based on eye tracking according to claim 8, characterized in that, In step S5, the local reconstruction is limited to adjusting only the visual expression parameters corresponding to the hierarchical deviation element group, without changing the display position and function type of the interface elements that are not identified as hierarchical deviation element groups.

10. The interface visual hierarchy optimization method based on eye tracking according to claim 9, characterized in that, In step S6, based on the fact that the number of hierarchical deviation element groups after optimization is less than the number of hierarchical deviation element groups before reconstruction, when the actual attention entry order of the target interface elements in the optimized actual perception hierarchical diagram is consistent with the expected attention entry order in the target visual hierarchical diagram, the optimized interface is determined as the target optimized interface.

Images