Presentation strategy evaluation and optimization method based on perception-judgment-decision link

Abstract

The invention relates to a presentation strategy evaluation and optimization method based on a perception-judgment-decision link, and solves the problems that in the existing intelligent auxiliary information situation interface design, the cognitive load is difficult to quantify, the task completion time evaluation is inaccurate, and the causal logic understanding lacks scientific indexes. According to the method, by constructing a multi-dimensional information entropy model, a task completion timeliness model and a causal definition model, the interface cognitive load, the task operation efficiency and the understanding degree of a user on task logic are quantified, then an optimal interface design scheme is screened through a multi-objective optimization method, the operation efficiency and the situation understanding ability of the user are improved, and the user experience is improved. And the information presentation effect of the intelligent auxiliary system is optimized. The method provided by the invention can effectively reduce cognitive load, shorten task completion time, enhance understanding of users on task causal relationships, and significantly improve usability and security of an intelligent auxiliary system interface, and has wide application value.

Description

Technical Field

[0001] This invention relates to the field of intelligent human-computer interaction and interface design, specifically to a method for evaluating and optimizing presentation strategies based on the perception-judgment-decision link. Background Technology

[0002] With the rapid development of information technology, the amount and complexity of information presented to users in modern intelligent systems have significantly increased, especially in fields such as command and control, intelligent cockpits, and medical diagnostic assistance, where users need to process large amounts of multi-source heterogeneous information. Interface design, as a crucial bridge connecting users and the system, directly impacts user work efficiency and task completion quality through its scientific rigor and rationality.

[0003] Traditional interface design relies heavily on the designer's experience and aesthetic principles, lacking effective quantification and control of user cognitive load, leading to information overload, user fatigue, and frequent operational errors. This is especially true in dynamic interfaces, where dynamic elements such as flashing icons and animations have an additional impact on users' visual attention and cognitive load, but existing design methods struggle to accurately assess the combined effect of these factors.

[0004] In addition, task completion time is an important indicator for evaluating the quality of interface design. However, traditional time evaluation models are mostly single-stage measurements, ignoring the hierarchical impact of the three key processes of perception, judgment and decision-making, and making it difficult to truly reflect the user's cognitive and operational processes in complex task environments.

[0005] More importantly, the user's understanding of the causal relationships within the task context and process significantly impacts the accuracy and safety of the operation. However, there is currently a lack of quantitative evaluation models for the causal relationships of interface information, resulting in interface design failing to adequately guarantee the user's clear understanding of the task logic.

[0006] Therefore, there is an urgent need for a scientific and effective interface evaluation and optimization method that can comprehensively quantify interface cognitive load, task completion timeliness, and causal clarity, and guide interface design accordingly to improve user experience and task execution effectiveness. Summary of the Invention

[0007] To address the technical problems in existing intelligent assisted information situational awareness interface designs, such as the difficulty in quantifying cognitive load, inaccurate assessment of task completion time, and the lack of scientific indicators for understanding causal logic, this invention proposes a method for evaluating and optimizing intelligent assisted information situational awareness enhancement strategies based on a perception-judgment-decision-making chain. By constructing a multi-dimensional information entropy model, a task completion timeliness model, and a causal clarity model, combined with multi-objective Pareto optimization, the scientific evaluation and optimization of interface design schemes can be achieved.

[0008] This invention is achieved through the following technical solutions:

[0009] A method for evaluating and optimizing presentation strategies based on the perception-judgment-decision link includes the following steps:

[0010] a) Use the information entropy model to divide the intelligent auxiliary interface into grids, calculate the symbol-weighted information content and dynamic interference terms of each grid of the interface, and combine color contrast and spatial density to correct the information entropy and obtain the effective information entropy value of the interface.

[0011] b) Construct a task completion time model based on the perception-judgment-decision (POD) link. Calculate the total task completion time by measuring perception time, judgment time, and decision time, combined with interference factors and task complexity.

[0012] c) Construct a causal network model among interface information elements and use a critical path identification algorithm to calculate the causal clarity index;

[0013] d) Based on effective information entropy, task completion timeliness, and causal clarity, a multi-objective optimization model is established, and a genetic algorithm is used to optimize the interface presentation strategy to obtain the Pareto front solution set;

[0014] e) Select the optimal presentation strategy from the Pareto frontier solution set based on actual usage requirements.

[0015] Furthermore, in the information entropy model, the weighted information content of symbols within the interface grid is summed by multiplying the weight coefficients with the element information entropy, and the dynamic interference term is calculated by combining the dynamic element weights with the dynamic interference coefficients.

[0016] Furthermore, in the task completion timeliness model, the perception time includes the basic perception time and the delay caused by dynamic interference and task complexity, the judgment time is adjusted according to the task complexity, and the decision time is corrected by combining the number of candidate solutions and information integrity indicators.

[0017] Furthermore, the causal clarity index is defined as the ratio of the sum of causal intensity of the critical path to the sum of causal intensity of the entire path, with a value ranging from 0 to 1.

[0018] Furthermore, the objective function of the multi-objective optimization model is to minimize the task completion time and maximize causal clarity, with the constraint that the effective information entropy does not exceed a preset threshold.

[0019] Furthermore, the genetic algorithm includes encoding interface design parameters, fitness function settings, selection, crossover, mutation, and termination conditions, and finally outputs the Pareto optimal solution set.

[0020] This invention constructs a multi-objective joint optimization intelligent evaluation framework by scientifically quantifying the complexity of interface information, user task completion efficiency, and depth of logical understanding. This method effectively reduces cognitive load, shortens task completion time, enhances users' understanding of causal relationships within tasks, and significantly improves the usability and security of intelligent assistance system interfaces, possessing broad application value. Attached Figure Description

[0021] Figure 1 Schematic diagram of the information entropy H model structure;

[0022] Figure 2 Flowchart of CT model for task completion time;

[0023] Figure 3 Schematic diagram of a causal network model;

[0024] Figure 4 Schematic diagram of Pareto frontier for multi-objective optimization;

[0025] Figure 5 Flowchart of this invention. Detailed Implementation

[0026] This invention is a presentation strategy evaluation and optimization method based on the perception-judgment-decision link. This method quantifies the interface cognitive load, task operation efficiency and user understanding of task logic by constructing a multi-dimensional information entropy model, a task completion time model and a causal clarity model. Then, it selects the optimal interface design scheme through a multi-objective optimization method to improve the user's operation efficiency and contextual understanding ability, and optimize the information presentation effect of the intelligent assistance system.

[0027] Figure 1 The overall structure of the information entropy model of this invention is illustrated. The interface is divided into several grid cells, each containing various information elements such as text, symbols, and icons. The arrows in the diagram represent the weighting process for information elements; the symbol-weighted information content and dynamic interference terms work together in the grid information content calculation module. The color contrast correction module receives interface color contrast data and performs non-linear adjustments to the information content. The spatial density correction module calculates the ratio of local to global information density, generates a correction factor, and finally summarizes the results to obtain the global cognitive load information entropy. Specifically:

[0028] 1) Information Entropy Model H

[0029] (1) Static symbol-weighted information content I:

[0030]

[0031] In the formula:

[0032] n ijIt represents the number of static element type j in grid i.

[0033] Average color contrast of static symbols in grid i.

[0034] H j Element type and information entropy.

[0035] w j : Element weight coefficient.

[0036] N: The total number of elements.

[0037] (2) Dynamic interference term Di:

[0038] Element-by-element modeling: For each dynamic element k (k = 1, 2, ..., m) i ) Calculate its interference terms independently, and then perform nonlinear accumulation.

[0039]

[0040] In the formula:

[0041] β k : The basic interference coefficient of the dynamic element k;

[0042] Color contrast of dynamic element k;

[0043] δ(f k ): Dynamic frequency correction factor, defined as:

[0044] f k : The frequency of change of dynamic element k (e.g., flicker frequency, unit: Hz);

[0045] f max The maximum frequency of dynamic elements in the interface.

[0046] (3) Total information content

[0047]

[0048] (4) Information density

[0049] Spatial density correction factor γ i

[0050] Definition: The amplification effect of local symbol density on cognitive load.

[0051]

[0052] Local density: The total number of symbols in mesh i (including dynamic elements);

[0053] Global density: The average symbol density across the entire interface.

[0054]

[0055] D i Information density of grid i (unit: bit / cm) 2 ).

[0056] (5) Weighted Information Entropy Formula

[0057]

[0058] in

[0059]

[0060] p i : Information density percentage of grid i;

[0061] H: Overall information entropy of the interface (unit: bit), reflecting the global cognitive load.

[0062] Figure 2 This illustrates the three-stage process of the Task Completion Time Efficiency Model (CT Model). First, the user observes through the interface, with the observation time including base time and delays caused by interference. Next, the user proceeds to the Orientation stage, where the Orientation time dynamically varies based on task complexity. Finally, the user makes a decision, with the decision time adjusted based on the number of candidate solutions and the completeness of information. This model reflects the hierarchical time consumption of the task completion process, revealing the impact of different factors on user operational efficiency. Specifically:

[0063] Based on the OODA loop theory, the task completion time is divided into three stages:

[0064] Perception time T P :

[0065] Including basic perception time μ P And delay of interference items:

[0066] T P =μ P +k 干扰 ·N 干扰项

[0067] Judgment time T J :

[0068] With task complexity D 任务复杂度 Proportional:

[0069] T J =μ J +k难度 ·D 任务复杂度

[0070] Decision time T D :

[0071] The base time is related to the logarithm of the number of candidate solutions and is also affected by information completeness K:

[0072]

[0073] Total task completion time:

[0074] CT=T P +T J +T D

[0075] Figure 3 This is a Causal Clarity Model (CC Model), which illustrates the causal network composed of interface information elements. Nodes represent different task objectives or key steps, edges represent causal relationships between nodes, and the thickness or color intensity of the edges reflects the strength of the causal relationship. The diagram identifies the critical path, i.e., the core task process link, and potentially interfering non-critical paths. This model is used to quantify the logical clarity of interface information and supports the calculation of causal clarity metrics. Specifically:

[0076] Construct a causal network of interface information elements, where nodes represent elements, edges represent causal relationships, and edge weights...

[0077] w is determined by expert scoring and data statistics.

[0078] Calculate the total strength of the critical path and total strength of all paths

[0079]

[0080] Causal clarity is defined as:

[0081]

[0082] At the same time, information entropy H is used as a constraint to avoid the causal network from becoming too chaotic.

[0083] Figure 4 This diagram illustrates the multi-objective optimization results of the present invention. The horizontal axis represents the task completion timeliness index CT (time, the smaller the better), and the vertical axis represents the causal clarity index CC (value range 0-1, the larger the better). The curves represent the Pareto front (the compromise solution between the two indices) for different interface design schemes. It is clearly visible in the figure that the schemes selected by the optimization algorithm achieve the optimal balance between CT and CC, satisfying the information entropy constraint H≤H. max The specific preferred method is as follows:

[0084] The optimization objectives are to minimize CT and maximize CC; the information entropy constraint H≤H is set. max The optimal interface design scheme is selected by using genetic algorithms or other heuristic algorithms combined with Pareto frontier analysis; ultimately, intelligent assistance and dynamic adjustment of interface design are achieved to improve user operation efficiency and contextual understanding.

[0085] The specific implementation steps of this invention are as follows:

[0086] Step 1: Interface Information Acquisition and Preprocessing

[0087] Collect interface design schemes for intelligent assistance systems and extract the type, quantity, and distribution of interface elements (text, icons, dynamic elements, etc.).

[0088] According to preset rules, the interface is divided into several grid units to facilitate subsequent information entropy calculation.

[0089] Step 2: Calculate the interface information entropy

[0090] For each element in the grid, assign a corresponding information entropy value and weight based on the element category.

[0091] The symbol-weighted information content and dynamic interference terms are calculated and corrected by combining color contrast.

[0092] The information density of each grid is calculated using a spatial density correction factor, and the probability distribution is normalized.

[0093] Calculate the overall effective information entropy of the interface and distinguish between useful and useless information.

[0094] Step 3: Build a task completion time model

[0095] Based on the task flow, determine the time baselines for the three stages of perception, judgment, and decision-making.

[0096] The time delay at each stage is calculated by combining the number of dynamic elements on the interface, the complexity of the task, and the number of candidate solutions.

[0097] Calculate the total task completion time as the timeliness target in the optimization.

[0098] Step 4: Establish causal network and calculate causal clarity

[0099] Based on the interface function modules and task steps, establish a causal network between information elements.

[0100] A critical path identification algorithm is used to extract critical paths and calculate causal strength.

[0101] Calculate the causal clarity index, which reflects the clarity of the causal relationship expressed on the interface.

[0102] Step 5: Multi-objective optimization modeling and algorithm solution

[0103] Construct a multi-objective optimization model with the goals of minimizing task completion time and maximizing causal clarity, constrained by effective information entropy.

[0104] Genetic algorithms are used for parameter encoding, fitness evaluation, selection, crossover, and mutation operations.

[0105] The Pareto optimal solution set is obtained by iterative solution.

[0106] Step 6: Solution Selection and Application

[0107] Based on actual operational needs and user feedback, the optimal interface presentation strategy is selected from the Pareto frontier solutions.

[0108] The selected solution will be applied to the intelligent assistance system for testing, verification, and further optimization.

[0109] Example:

[0110] The steps of this invention include:

[0111] 1) Construction and calculation of the interface information entropy model

[0112] (1) Grid generation and element classification

[0113] The intelligent assistance interface to be evaluated is divided into several grid units according to regional function and visual density (e.g., Figure 1 (As shown). Each grid contains multiple types of elements, mainly categorized as follows:

[0114] Text-based elements (such as instructions and labels)

[0115] Icon elements (standard symbols, status indicators)

[0116] Dynamic elements (blinking icons, animations)

[0117] The number and type of elements are automatically extracted using the interface parsing tool.

[0118] (2) Information entropy and weight setting

[0119] For each type of element, the information entropy value H is determined based on its complexity and cognitive load impact. j and weight w j .For example:

[0120] The text has high information entropy and high weight (e.g., w). 文字 =0.5)

[0121] The icon has moderate information entropy and appropriate weight (e.g., w).图标 =0.3)

[0122] Dynamic elements have higher weights, and a dynamic interference coefficient β (e.g., β = 0.7) is introduced.

[0123] (3) Color contrast calculation

[0124] Color contrast C i Calculate using Weber's formula:

[0125]

[0126] Where L represents the brightness value, ensuring that visual stimuli are effectively quantified.

[0127] (4) Calculation process

[0128] Static symbol correction information content: It is necessary to calculate the weighted information content of each static element and multiply it by the color contrast correction factor.

[0129] Dynamic interference correction: Calculate the interference term for each dynamic element one by one, taking into account its color contrast, basic interference coefficient and frequency correction factor, and then sum them up.

[0130] Spatial density correction factor: The nonlinear correction factor is calculated based on the local density and the global average density.

[0131] Total information content and information density: The static and dynamic components are added together, spatial density correction is applied, and finally the information density of each grid is obtained.

[0132] Overall information entropy: Calculated based on the weighted information entropy ratio of information density.

[0133] 2) Application of the task completion timeliness model

[0134] (1) Parameter determination

[0135] Baseline time parameters were determined experimentally.

[0136] Such as basic sensing time μ P = 600 milliseconds

[0137] Such as the basic judgment time μ J =800 milliseconds

[0138] Such as the basic decision-making time μ D =1000 milliseconds

[0139] The delay coefficient of interference items, the weight of task complexity, and the coefficient of the number of options were obtained through user testing and statistical fitting.

[0140] (2) Task decomposition and time estimation

[0141] Based on the specific task flow, the perception, judgment, and decision-making stages are broken down. The number of operational elements, interference items, and options corresponding to each stage are counted, and the overall completion time (CT) is calculated.

[0142] For example, if a user operation contains 3 dynamic interference elements, the task complexity metric is 0.7 (out of 1), the number of candidate solutions is 4, and the information integrity index K = 0.9, then substitute these values ​​into the formula to obtain the time estimate.

[0143] 3) Implementation of the causal clarity model

[0144] (1) Causal network construction

[0145] Based on the task scenario, key task steps and their causal relationships are identified, and edge weights are determined by combining expert scoring with system log statistics.

[0146] (2) Critical path identification

[0147] The core process path set of the task is extracted using graph theory shortest path algorithm or critical path method.

[0148] (3) Causal clarity calculation

[0149] Calculate the critical path causal total strength and the total causal total strength of the entire path, calculate the ratio to obtain the CC, and evaluate the clarity of the causal expression at the interface.

[0150] 4) Multi-objective optimization and interface selection

[0151] (1) Optimize target setting

[0152] Objective function:

[0153] Minimize task completion time (CT)

[0154] Maximize Causal Clarity (CC)

[0155] Constraints:

[0156] Effective information entropy H eff ≤H max

[0157] (2) Optimization method selection

[0158] A genetic algorithm is used as the optimization framework, with individual codes representing interface design parameters (such as element layout, number of dynamic elements, and color scheme). New solutions are generated through crossover and mutation.

[0159] (3) Results Analysis and Screening

[0160] Through algorithmic iteration, a set of Pareto front solutions is obtained, and designers select the optimal solution based on actual usage requirements.

[0161] 5) Application Examples

[0162] Taking the intelligent cockpit interface as an example, the method of this invention is applied to compare three different interface design schemes:

[0163] Option A: Information-intensive, with many dynamic elements

[0164] Option B: Moderate information, appropriate amount of dynamic elements

[0165] Option C: Simplified information, fewer dynamic elements

[0166] After calculations using the information entropy model, CT model, and CC model, and multi-objective optimization, the results show that Scheme B maintains high causal clarity while achieving a lower task completion time and moderate cognitive load, and is recommended as the final design scheme.

[0167] Figure 5 The overall implementation process of the method of this invention is described, and the functions of each module in the figure are explained:

[0168] 1) Information Entropy Model Module: Receives the interface design scheme, divides the interface into grids, calculates the symbol-weighted information content and dynamic interference terms, and corrects the information entropy by combining color contrast and spatial density to quantify the information load of the interface.

[0169] 2) Task completion time model module: Based on the perception-judgment-decision link, estimate the time required for different stages, and calculate the total task completion time by combining dynamic elements of the interface and task complexity.

[0170] 3) Causal Network Model Module: Constructs a causal relationship network among interface information elements, uses a critical path identification algorithm to calculate causal clarity, and reflects the clarity of the causal logic expression in the interface.

[0171] 4) Multi-objective optimization model module: Establish a multi-objective optimization model with the objectives of minimizing task completion time and maximizing causal clarity, and with information entropy as the constraint condition, and use a genetic algorithm to solve it.

[0172] 5) Presentation strategy optimization and output: Select the optimal interface presentation strategy from the Pareto front solution set obtained by the genetic algorithm, optimize the interface of the intelligent assistance system, and improve cognitive efficiency and task completion quality.

[0173] 6) User feedback and continuous optimization: Implement the optimized interface design, collect user operation feedback, and carry out continuous improvement and iterative optimization.

[0174] Specific procedures:

[0175] Input phase

[0176] - Receive interface design data and task requirement information.

[0177] Data preprocessing

[0178] - Grid division and element classification.

[0179] Information entropy calculation

[0180] - Calculate the symbol-weighted information content.

[0181] - Calculate the dynamic disturbance term.

[0182] - Color contrast correction.

[0183] - Calculate the effective information entropy.

[0184] Task time estimation

[0185] - Calculation of perception, judgment, and decision-making time.

[0186] -Calculation of overall task completion time.

[0187] Causal network analysis

[0188] - Causal network construction.

[0189] - Critical path identification.

[0190] - Causal clarity calculation.

[0191] Multi-objective optimization

[0192] - Definition of the objective function.

[0193] - Solve using a genetic algorithm.

[0194] - Obtain the Pareto solution set.

[0195] Output and Solution Selection

[0196] - Optimal interface presentation strategy. - Application and verification.

Claims

1. A method for evaluating and optimizing presentation strategies based on a perception-judgment-decision link, characterized in that: Includes the following steps: a) Use the information entropy model to divide the intelligent auxiliary interface into grids, calculate the symbol-weighted information content and dynamic interference terms of each grid of the interface, and combine color contrast and spatial density to correct the information entropy and obtain the effective information entropy value of the interface. b) Construct a task completion time model based on the perception-judgment-decision (POD) link. Calculate the total task completion time by measuring perception time, judgment time, and decision time, combined with interference factors and task complexity. c) Construct a causal network model among interface information elements and use a critical path identification algorithm to calculate the causal clarity index; d) Based on effective information entropy, task completion timeliness, and causal clarity, a multi-objective optimization model is established, and a genetic algorithm is used to optimize the interface presentation strategy to obtain the Pareto front solution set; e) Select the optimal presentation strategy from the Pareto frontier solution set based on actual usage requirements.

2. The method for evaluating and optimizing a presentation strategy based on a perception-judgment-decision link according to claim 1, characterized in that: In the information entropy model, the weighted information content of symbols within the interface grid is summed by multiplying the weight coefficients with the element information entropy, and the dynamic interference term is calculated by combining the dynamic element weights with the dynamic interference coefficients.

3. The method for evaluating and optimizing a presentation strategy based on a perception-judgment-decision link according to claim 1, characterized in that: In the task completion timeliness model, the perception time includes the basic perception time and the delay caused by dynamic interference and task complexity. The judgment time is adjusted according to the task complexity, and the decision time is corrected by combining the number of candidate solutions and information integrity indicators.

4. The method for evaluating and optimizing a presentation strategy based on a perception-judgment-decision link according to claim 1, characterized in that: The causal clarity index is defined as the ratio of the sum of causal intensity of the critical path to the sum of causal intensity of the entire path, and its value ranges from 0 to 1.

5. The method for evaluating and optimizing a presentation strategy based on a perception-judgment-decision link according to claim 1, characterized in that: The objective function of the multi-objective optimization model is to minimize the task completion time and maximize causal clarity, with the constraint that the effective information entropy does not exceed a preset threshold.

6. The method for evaluating and optimizing a presentation strategy based on a perception-judgment-decision link according to claim 1, characterized in that: The genetic algorithm includes encoding interface design parameters, fitness function settings, selection, crossover, mutation, and termination conditions, and finally outputs the Pareto optimal solution set.

Images