Human-computer interaction degree evaluation method based on bidirectional information value

Abstract

The invention belongs to the technical field of man-machine interaction in artificial intelligence, and particularly relates to a man-machine interaction degree evaluation method based on bidirectional information value, which comprises the following steps: calculating the man-machine interaction degree of a single man to a machine; calculating the man-machine interaction degree of a single machine to a person; calculating man-machine interaction degree parameters; and calculating and outputting the total human-computer interaction degree, and evaluating the efficiency and effect of human-computer interaction. According to the method, the problems of incomplete man-machine interaction degree calculation and lack of objective quantitative standards in existing unmanned aerial vehicle autonomy evaluation are solved. A two-way information value concept is introduced for the first time, and an interaction scene is fully covered; subjective expert experience is quantified through an analytic hierarchy process, and evaluation objectivity is improved; and in combination with time dynamics and content effectiveness, the model engineering applicability is enhanced. The accuracy and credibility of unmanned aerial vehicle autonomy evaluation can be remarkably improved, and a scientific basis is provided for command decision making.

Description

Technical Field

[0001] This invention belongs to the field of human-computer interaction technology in artificial intelligence, specifically a method for evaluating human-computer interaction based on two-way information value. Background Technology

[0002] Currently, in the classification of UAV autonomy levels, human-machine interaction is a core sub-indicator of autonomy, directly affecting whether the system requires frequent human intervention. While current mainstream assessment systems (such as the U.S. Department of Defense's Autonomy Rating Scale (ALFUS) use interaction frequency as a grading standard, they do not deeply quantify the quality and timeliness of the interaction content. For example, with the same number of interactions, high-value commands and inefficient, lengthy operations are treated equally, leading to distorted assessments. Furthermore, existing technologies often focus on one-way interactions (such as operator command response speed), neglecting the role of proactive UAV feedback (such as situational awareness sharing and fault assistance) in improving overall autonomy. This limitation stems from two technical bottlenecks: first, the lack of a unified standard for evaluating interaction value means that information usefulness often relies on subjective qualitative descriptions, making it difficult to integrate into mathematical models; second, the time factor has not been systematically modeled, failing to reflect the value difference between early critical commands and later redundant operations. More seriously, in certain environments, human-machine interaction is highly dynamic, making traditional static weighting methods (such as fixed-coefficient weighting) difficult to adapt to real-time changes. These problems render the existing human-computer interaction evaluation results unreliable, severely restricting the accurate optimization of UAV autonomous capabilities and the efficiency of command and decision-making.

[0003] The core evaluation indicators for drone autonomy include human-machine interaction, complexity, and completion rate. Among these, human-machine interaction directly reflects the collaborative effectiveness between the system and the operator. Existing technologies have two major drawbacks: first, the calculation of interaction rate relies heavily on unidirectional indicators (such as focusing only on operator commands or drone feedback), neglecting the collaborative impact of two-way information flow; second, there is a lack of dynamic quantification of information timeliness, content value, and interactive initiative, leading to one-sided evaluation results. Summary of the Invention

[0004] To address the technical problems existing in the prior art, this invention proposes a human-computer interaction evaluation method based on two-way information value, comprising the following steps: Step 1: Calculate the human-computer interaction degree of a single human-to-machine interaction; Step 2: Calculate the human-computer interaction degree of a single machine-to-human interaction; Step 3: Calculate the human-computer interaction parameters; Step 4: Calculate and output the total human-computer interaction score, and evaluate the efficiency and effectiveness of human-computer interaction.

[0005] Furthermore, step one specifically includes: 101. Accelerating information exchange through computation and the criticality of control time Measuring the time value of information provided by humans to machines Among them, the earlyness of information interaction for:

[0006] in, Indicates the end time of the task. Indicates the time when the information was sent. Indicates the total duration of the entire task; For transient operation modes, the criticality of the control moment. for: ; For continuous operation mode, the criticality of control timing. for:

[0007] in, Indicates the moment when control ends. Indicates the moment when control begins. Indicates the time when the task begins. This indicates the moment the task ends; 102. Evaluate the content value of information provided by machines to humans by calculating information accuracy, information completeness, and information suitability. ,in, ; ; ; 103. Evaluate the human operator's control efficiency of the machine by calculating one or more of the following: effective control time ratio, operational effectiveness, and response efficiency. Among them, the proportion of effective control time for:

[0008] In the formula, For effective human control time, This refers to the total execution time of the task; Operational effectiveness for: ; Response efficiency for: ; 104. The interactive initiative factor of human-to-machine computation : ; 105. Calculate the human-computer interaction degree of a single human-to-machine interaction: For the first... Secondary human-computer interaction, human-to-machine human-computer interaction degree for:

[0009] In the formula, Indicates the weighting coefficient; Indicates the first In this interaction, humans provide information content value to machines; Indicates the first The time value of information provided by humans to machines in this interaction; Indicates the first The proportion of control time in each interaction; This represents the operator's interactive initiative factor.

[0010] Furthermore, step two specifically includes: 201. Accelerating information exchange through computation Measure the first The time value of information provided by the machine to the human in this interaction ; 202. Calculating the accuracy of information Information completeness Information Adaptability and information comprehensibility One or more of the evaluations in In this interaction, the value of the information provided by the machine to the human is... ; 203. The active factors of computer-to-human interaction : ; 204. Calculate the overall human-computer interaction degree of a single machine-to-human interaction, for the first... Secondary human-computer interaction, machine-to-human human-computer interaction degree for:

[0011] In the formula, Indicates the weighting coefficient; Indicates the first In this interaction, the machine provides the content value of information to the human; Indicates the first The time value of information provided by the machine to the human during this interaction; This represents the initiative factor in machine interaction.

[0012] Furthermore, in step 202, the comprehensibility of the evaluation information is assessed. Use one of the following two methods: Qualitative assessment: Through questionnaires and interviews, we explored the operators' understanding and attitudes toward the trust formation mechanism. The scores were assigned based on the degree of comprehensibility, and then normalized. Quantitative calculation: The formula is: .

[0013] Furthermore, step three specifically includes: 301. Construct a hierarchical model, in which the target layer is the overall human-computer interaction degree, the criterion layer includes human-to-computer interaction degree and machine-to-human interaction degree, and the solution layer includes each human-computer interaction event; 302. Construct the judgment matrix; 303. Calculate the weight vector; 304. Perform a consistency check.

[0014] Furthermore, step 302 specifically includes: 302a. Construct the criterion layer judgment matrix A: For the human-to-computer interaction degree and the machine-to-human interaction degree in the criterion layer, pairwise comparisons are performed according to the "1-5 scale method" to construct the criterion layer judgment matrix A; 302b. Construct the human-to-computer interaction degree judgment matrix A h : For the n specific influencing factors of human-to-computer interaction, pairwise comparisons are performed using the "1-9 scale method" to construct a judgment matrix A. h Judgment matrix A h The element in is denoted as a ij ; 302c. Determine the weight of machine-to-human interaction through a scoring method.

[0015] Furthermore, step 303 specifically includes: For the judgment matrix Normalize by columns to obtain the normalized matrix. ; For normalized matrix The weight vector is obtained by summing the elements in each row and dividing by the number of elements. .

[0016] Furthermore, in step 303, the specific method for normalization is: to normalize the judgment matrix. Each element Divide by the sum of the elements in its column to obtain the normalized matrix. elements ,Right now:

[0017] In the formula, To determine the order of a matrix.

[0018] Furthermore, step 304 specifically includes: Calculate the judgment matrix The largest eigenvalue The calculation formula is as follows:

[0019] In the formula, Represents the judgment matrix With weight vector The first new vector obtained by multiplication One component; Represents the weight vector The One component; Indicates the order of the judgment matrix; Calculate the consistency index The calculation formula is as follows: ; According to the judgment matrix order Sure The possible values ​​of ; Calculate the consistency ratio The calculation formula is as follows:

[0020] In the formula, As a consistency indicator, The average random consistency index; when At that time, it is assumed that the judgment matrix has consistency. That is, the weight; otherwise, modify the judgment matrix until the consistency check is passed.

[0021] Furthermore, in step four, the total human-computer interaction degree... for:

[0022] In the formula, and These represent the number of human-to-computer interactions and the number of computer-to-human interactions, respectively. These are adjustable parameters.

[0023] This invention outputs the total human-computer interaction parameter through multi-dimensional weighted fusion. It also achieves comprehensive evaluation coverage of the entire interaction chain through bidirectional modeling, avoiding the one-sidedness of traditional methods. Furthermore, the combination of the Analytic Hierarchy Process (AHP) and consistency checks preserves expert knowledge while reducing subjective bias. In addition, this invention introduces a time value algorithm to enhance the model's adaptability to real-time tasks. Attached Figure Description

[0024] Figure 1 A flowchart for calculating human-computer interaction parameters; Figure 2 A hierarchical model diagram for human-computer interaction evaluation. Detailed Implementation

[0025] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. 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 should fall within the scope of protection of the present invention.

[0026] This invention constructs a comprehensive and objective completion method model. The "human" mentioned in this invention refers to the operator, not the machine's task object. The human-machine interaction mentioned in this invention mainly refers to the interaction between the operator and the unmanned system in sharing and supplementing task information, excluding the interaction between the unmanned system and the task object.

[0027] In one specific embodiment, the evaluation is carried out through the following four steps: calculating the human-computer interaction degree of a single human-to-machine interaction, calculating the human-computer interaction degree of a single machine-to-human interaction, calculating the human-computer interaction degree parameter, and calculating the total human-computer interaction degree.

[0028] Step 1: Calculate the human-computer interaction degree of a single human-to-machine interaction.

[0029] 101. Calculating the Time Value of Information (Human-to-Machine Interaction) The time value of information provided by a person to a machine in the i-th interaction represents the timeliness of the information provided. It is expressed as follows. It is mainly measured by two indicators: the timeliness of information interaction and the criticality of the control moment.

[0030] (1) Information exchange advance When the help information is feature-based or judgment-based, the time value of the information depends on the brief interaction ante. The brief interaction ante reflects how far ahead the information is provided relative to the optimal time. The higher the brief interaction ante, the higher the time value of the information, and vice versa. It can be represented as:

[0031] in, Indicates the end time of the task. Indicates the time when the information was sent. Indicates the total duration of the entire task. Brief information advance rate. The range is [0,1]. The earlier the information is sent, the closer the value is to 1. If the information is sent at the beginning of the task, the value is 1. If the information is sent after the task ends, the value is 0, indicating that the information has been sent invalidally.

[0032] (2) Controlling the criticality of the moment When the help information is control-related information, the time value of the information depends on the criticality of the control moment. The criticality of the control moment reflects the importance of the moment when the control action occurs to the completion of the task. The higher the criticality of the control moment, the higher the time value of the information, and vice versa.

[0033] For transient operation mode It can be represented as:

[0034] For continuous operation mode, It can be represented as:

[0035] in, Indicates the moment when control ends. Indicates the moment when control begins. This indicates the moment when the task begins (or the time when a subtask needs to be performed). This indicates the moment the task ends. Control the criticality of the timing. The range is [0,1], and the closer it is to 1, the more precise the control time (or time period) is.

[0036] 102. Calculating the value of information content (human-to-machine) This reflects the degree to which the information provided by machines meets the machine's task requirements. The content value of information provided by humans to machines in this interaction is represented by... The accuracy, completeness, and adaptability of the indicators can be reflected in aspects such as accuracy, completeness, and adaptability. For different unmanned systems and application scenarios, one or more of the three indicators can be selected. For example, for situations requiring a large amount of information with rich content, information completeness can be selected; for open-ended, question-and-answer interaction modes, information adaptability can be selected. When multiple indicators are selected, a comprehensive value can be obtained by weighting them according to their importance.

[0037] (1) Information accuracy, including the degree to which information matches facts, logical consistency, and error control. The calculation formula is as follows:

[0038] (2) Completeness: The comprehensiveness of the information coverage, avoiding the omission of important information. The calculation formula is as follows:

[0039] (3) Adaptability: The degree to which the information provided by the system matches the application scenario, problem, etc., and the extent to which it meets information requirements. Its calculation formula is:

[0040] The range of the three indicators mentioned above is [0,1]. The closer the value is to 1, the higher the value of the information content.

[0041] 103. Calculate control efficiency Human-machine control efficiency This refers to a comprehensive indicator that measures the relationship between the time resources required for a user to operate a machine and the results achieved in completing a specific task. It reflects the usability of the human-computer interface and the system performance. Its core evaluation dimensions typically include indicators such as the effective control time ratio, operational effectiveness, and response efficiency. For different unmanned systems and application scenarios, one or more of these three indicators can be selected. For example, if the human intervention mode and ratio of certain unmanned systems are relatively fixed, the "effective control time ratio" indicator may not be used, and more emphasis can be placed on operational effectiveness and response efficiency. When multiple indicators are selected, a comprehensive value can be obtained through a weighted approach based on importance.

[0042] (1) The proportion of effective control time reflects the degree of human intervention, which can be measured by the effective human control time. Total task execution time The ratio is expressed as:

[0043] The value of this indicator is in the range of [0,1]. The closer it is to 0, the lower the system's dependence on humans.

[0044] (2) Operational effectiveness refers to the proportion of invalid operations during machine operation. Its calculation formula is:

[0045] The value of this indicator is in the range of [0,1], and the closer it is to 1, the fewer invalid operations there are.

[0046] (3) Response efficiency refers to whether operators can provide timely assistance or feedback when the system requires intervention. The calculation formula is:

[0047] The value of this indicator is in the range of [0,∞], and the closer it is to 0, the higher the response efficiency.

[0048] 104. Calculate the interaction initiative factor (human-to-machine) The interaction initiative factor reflects whether a single intervention is triggered by a request for help from the unmanned system or initiated by the operator's subjective will. Here, it primarily refers to the initiative of humans in providing assistance to machines. It can be calculated using the following formula:

[0049] The value ranges from [0,1], and the closer it is to 1, the stronger the initiative.

[0050] 105. Calculate the degree of a single human-computer interaction (human-to-machine).

[0051] For the Secondary human-computer interaction, defining the degree of human-to-computer interaction. for:

[0052] In the formula, This represents the weighting coefficient, used to adjust the degree of influence of each factor on the interaction degree; Indicates the first In this interaction, humans provide information content value to machines; Indicates the first The time value of information provided by humans to machines in this interaction; Indicates the first The proportion of control time in each interaction; This represents the operator's interactive initiative factor.

[0053] Step 2. Calculate the human-computer interaction degree of a single machine-to-human interaction.

[0054] 201. Calculating the Time Value of Information (Machine-to-Human) The time value of information provided by the machine to the human in the j-th interaction represents the timeliness of the information provided by the machine to the human. This is used to represent the degree of information interaction. Since machine-provided information to humans typically does not include the human operational aspect, compared to human-to-machine interaction, this section primarily calculates the information interaction primacy. For the specific calculation formula, please refer to step 101.

[0055] 202. Calculate the value of information content (machine-to-human) This reflects the degree to which the information provided by the machine meets the needs of the operators. In this interaction, the content value of the information provided by the machine to the human is represented as follows: Similar to the value of information provided by machines to humans, this can also be assessed based on accuracy. Completeness Adaptability In addition to reflecting these aspects, the comprehensibility of information should also be improved. This reflects the degree to which operators can understand the system's decision-making process and logic.

[0056] Methods for evaluating comprehensibility can employ qualitative judgment, using questionnaires, interviews, etc., to uncover operators' understanding and attitudes towards trust formation mechanisms. Scores are then assigned based on higher comprehensibility, and normalized. Alternatively, quantitative calculation methods can be used.

[0057] During implementation, one or more of the four indicators can be selected selectively.

[0058] 203. Calculate the interactive active factor (machine-to-human) This refers to the extent to which a system proactively detects and anticipates problems and provides information to people. It can be calculated using the following formula:

[0059] The value ranges from [0,1], and the closer it is to 1, the stronger the initiative.

[0060] 204. Calculate the degree of a single human-computer interaction (machine-to-human) interaction. For the Secondary human-computer interaction, defining the degree of machine-to-human interaction. for:

[0061] In the formula, This represents the weighting coefficient, used to adjust the degree of influence of each factor on the interaction degree; Indicates the first In this interaction, the machine provides the content value of information to the human; Indicates the first The time value of information provided by the machine to the human during this interaction; This represents the initiative factor in machine interaction.

[0062] Step 3: Calculate the human-computer interaction parameters.

[0063] like Figure 1 As shown, the process is implemented in four steps: establishing a hierarchical structure model, constructing a judgment matrix, calculating weight vectors, and performing consistency checks.

[0064] 301. Constructing a hierarchical model First, a three-level hierarchical structure model for evaluating human-computer interaction is established, and the structure of the model is as follows: Figure 2 As shown.

[0065] (1) Target layer: Located at the top of the model, it represents the final goal of the evaluation, namely "overall human-computer interaction degree".

[0066] (2) Criterion layer: Located in the middle layer of the model, it contains two parallel and independent evaluation criteria: Human-to-machine interaction: used to evaluate the efficiency and effectiveness of users issuing commands or information to machines.

[0067] Machine-to-human interaction: Used to evaluate the efficiency and effectiveness of machines in providing feedback to users.

[0068] (3) Solution layer: Located at the bottom of the model, it consists of all the specific objects to be evaluated, i.e., each "human-computer interaction event". Each specific event must be evaluated by the above two criteria at the same time.

[0069] 302. Construct the judgment matrix In this step, based on the differences in the magnitude and importance of the comparison factors, two scaling methods are used for quantitative judgment: the 1-9 scaling method is shown in Table 1, and the 1-5 scaling method is shown in Table 2.

[0070] Table 1. Explanation of the meaning of the 1-9 scale method

[0071] Table 2. Explanation of the meaning of the 1-5 scaling method

[0072] Based on the scaling method described above, the specific implementation of this step is as follows: 302a. Construct the criterion layer judgment matrix A.

[0073] Experts were invited to conduct pairwise comparisons of the two factors at the criterion layer—"human-to-computer interaction" and "computer-to-human interaction"—based on their relative importance to achieving the overall goal, using the 1-5 scale method shown in Table 2. Based on the comparison results, a 2x2 criterion layer judgment matrix, denoted as A, was constructed.

[0074] 302b. Construct the human-to-computer interaction degree judgment matrix A h .

[0075] For the n specific influencing factors under the "human-to-computer interaction degree" criterion, experts were invited to conduct pairwise comparisons of these factors using the 1-9 scale method shown in Table 1. Based on the comparison results, an n x n judgment matrix was constructed, denoted as A.h Matrix A h The element in is denoted as a ij , representing the importance scale value of influencing factor i relative to influencing factor j under the "human-to-computer interaction degree" criterion. This matrix also satisfies basic properties such as reciprocity.

[0076] 302c. Determine the weights of machine-to-human interaction.

[0077] In this embodiment, only one key influencing factor (such as "usefulness of help information") is set under the "machine-to-human interaction degree" criterion. Its weight is determined by a direct expert rating method: K experts (K≥3) are invited to independently rate the score (e.g., on a percentage scale), and the rating of the kth expert is s. k Then the weight parameter of this factor The arithmetic mean of all expert ratings.

[0078] 303. Calculate the weight vector For the judgment matrix Normalize by columns to obtain the normalized matrix. The purpose of normalization is to convert elements with different dimensions into dimensionless relative importance values, facilitating comparison and calculation. The specific method of normalization is to transform the judgment matrix... Each element Divide by the sum of the elements in its column to obtain the normalized matrix. elements ,Right now:

[0079] In the formula, To determine the order of a matrix, perform a normalized matrix... The weight vector is obtained by summing the elements in each row and dividing by the number of elements. .

[0080] 304. Consistency Check The purpose of consistency testing is to determine whether there are logical contradictions or random errors in the subjective judgments of experts, thereby ensuring the credibility and validity of the evaluation results. The basic principle of the test is that for a completely consistent judgment matrix, its largest eigenvalue is equal to the order of the matrix; conversely, the more inconsistent the judgment matrices are, the greater the deviation between the largest eigenvalue and the order. Therefore, the consistency of the judgment matrix can be measured by calculating the deviation between the largest eigenvalue and the order.

[0081] ① Calculate the judgment matrix The largest eigenvalue The calculation formula is as follows:

[0082] In the formula, Representation matrix With weight vector The first new vector obtained by multiplication One component; Represents the weight vector The One component; This indicates the order of the judgment matrix.

[0083] ② Calculate the consistency index The calculation formula is as follows:

[0084] ③ Find the average random consistency index . Consistency index is applied to a large number of randomly generated judgment matrices. The calculated average value reflects the average degree of inconsistency of the judgment matrix. The value of and the order of the matrix related, By 9 o'clock The values ​​are shown in Table 3.

[0085] Table 3. Mapping of RI values ​​to matrix order n

[0086] According to the judgment matrix order Determine by looking up the table The value of .

[0087] ④ Calculate the consistency ratio The calculation formula is as follows:

[0088] In the formula, As a consistency indicator, As the average random consistency index, when At that time, it is considered that the judgment matrix has satisfactory consistency. This is the weight; otherwise, the judgment matrix needs to be modified until the consistency check is passed.

[0089] Step 4: Calculate the total human-computer interaction degree.

[0090] Taking all factors into consideration Secondary human-computer interaction, defining the overall human-computer interaction degree. for:

[0091] In the formula, and These represent the number of human-to-computer interactions and the number of computer-to-human interactions, respectively. This is an adjustable parameter used to control the amplification effect of the number of interactions on the overall interactivity. The overall human-computer interactivity ranges from [0,1), and the closer it is to 1, the higher the interaction efficiency and the better the interaction effect.

[0092] In the technical solution of this invention: 1) Human-computer interaction is decomposed into two dimensions: "human-to-machine" and "machine-to-human," and quantitative sub-models are established for each. The degree of human-to-machine interaction is calculated by comprehensively considering the usefulness of the help information, time value (reflecting the timeliness of instructions), proportion of control time (measuring operational load), and interaction initiative factor (distinguishing between active intervention and passive response); the degree of machine-to-human interaction is evaluated based on the usefulness of the help information and the initiative factor.

[0093] 2) The AHP method is used to integrate expert experience and objective data to construct a three-layer hierarchical structure (target layer - criterion layer - solution layer). The weight of the indicators is quantified by the judgment matrix and subjective contradictions are eliminated through consistency verification.

[0094] 3) Propose a dynamic decay algorithm for the time value of information to ensure that the contribution of high-value instructions in the early stage is not diluted by inefficient interactions in the later stage.

[0095] This invention achieves its goal of outputting a total human-computer interaction parameter through multi-dimensional weighted fusion. Its technical effects are reflected in three aspects: First, bidirectional modeling enables the evaluation to cover the entire interaction chain, avoiding the one-sidedness of traditional methods; second, the combination of AHP and consistency testing preserves expert knowledge while reducing subjective bias; and third, the introduction of the time value algorithm enhances the model's adaptability to real-time tasks.

[0096] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. A method for evaluating human-computer interaction based on two-way information value, characterized in that, Includes the following steps: Step 1: Calculate the human-computer interaction degree of a single human-to-machine interaction; Step 2: Calculate the human-computer interaction degree of a single machine-to-human interaction; Step 3: Calculate the human-computer interaction parameters; Step 4: Calculate and output the total human-computer interaction score, and evaluate the efficiency and effectiveness of human-computer interaction.

2. The method according to claim 1, characterized in that, Step one specifically includes:

101. Accelerating information exchange through computation and the criticality of control time Measuring the time value of information provided by humans to machines Among them, the earlyness of information interaction for: in, Indicates the end time of the task. Indicates the time when the information was sent. Indicates the total duration of the entire task; For transient operation modes, the criticality of control timing. for: ； For continuous operation mode, the criticality of control timing. for: in, Indicates the moment when control ends. Indicates the moment when control begins. Indicates the moment the task begins. This indicates the moment the task ends; 102. Evaluate the content value of information provided by machines to humans by calculating information accuracy, information completeness, and information suitability. ,in, ； ； ； 103. Evaluate the operator's control efficiency over the machine by calculating one or more of the following: effective control time ratio, operational effectiveness, and response efficiency. Among them, the proportion of effective control time for: In the formula, For effective human control time, This refers to the total execution time of the task; Operational effectiveness for: ； Response efficiency for: ； 104. The interactive initiative factor of human-to-machine computation : ； 105. Calculate the human-computer interaction degree of a single human-to-machine interaction: For the first... Secondary human-computer interaction, human-to-machine human-computer interaction degree for: In the formula, Indicates the weighting coefficient; Indicates the first In this interaction, humans provide information content value to machines; Indicates the first The time value of information provided by humans to machines in this interaction; Indicates the first The proportion of control time in each interaction; This represents the operator's interactive initiative factor.

3. The method according to claim 1, characterized in that, Step two specifically includes:

201. Accelerating information exchange through computation Measure the first The time value of information provided by the machine to the human in this interaction ; 202. Calculating the accuracy of information Information completeness Information Adaptability and information comprehensibility One or more of the evaluations in In this interaction, the value of the information provided by the machine to the human is... ; 203. The active factors of computer-to-human interaction : ； 204. Calculate the overall human-computer interaction degree of a single machine-to-human interaction, for the first... Secondary human-computer interaction, machine-to-human human-computer interaction degree for: In the formula, Indicates the weighting coefficient; Indicates the first In this interaction, the machine provides the content value of information to the human; Indicates the first The time value of information provided by the machine to the human during this interaction; This represents the initiative factor in machine interaction.

4. The method according to claim 3, characterized in that, In step 202, the comprehensibility of the evaluation information is assessed. Use one of the following two methods: Qualitative assessment: Through questionnaires and interviews, we explored the operators' understanding and attitudes toward the trust formation mechanism. The scores were assigned based on the degree of comprehensibility, and then normalized. Quantitative calculation: The formula is: 。 5. The method according to claim 1, characterized in that, Step three specifically includes:

301. Construct a hierarchical model, in which the target layer is the overall human-computer interaction degree, the criterion layer includes human-to-computer interaction degree and machine-to-human interaction degree, and the solution layer includes each human-computer interaction event; 302. Construct the judgment matrix; 303. Calculate the weight vector; 304. Perform a consistency check.

6. The method according to claim 5, characterized in that, Step 302 specifically includes: 302a. Construct the criterion layer judgment matrix A: For the human-to-computer interaction degree and the machine-to-human interaction degree in the criterion layer, pairwise comparisons are performed according to the "1-5 scale method" to construct the criterion layer judgment matrix A; 302b. Construct the human-to-computer interaction degree judgment matrix A h : For the n specific influencing factors of human-to-computer interaction, pairwise comparisons are performed using the "1-9 scale method" to construct a judgment matrix A. h Judgment matrix A h The element in is denoted as a ij ; 302c. Determine the weight of machine-to-human interaction through a scoring method.

7. The method according to claim 6, characterized in that, Step 303 specifically includes: For the judgment matrix Normalize by columns to obtain the normalized matrix. ; For normalized matrix The weight vector is obtained by summing the elements in each row and dividing by the number of elements. .

8. The method according to claim 7, characterized in that, In step 303, the specific method of normalization is: to normalize the judgment matrix. Each element Divide by the sum of the elements in its column to obtain the normalized matrix. elements ,Right now: In the formula, To determine the order of a matrix.

9. The method according to claim 7, characterized in that, Step 304 specifically includes: Calculate the judgment matrix The largest eigenvalue The calculation formula is as follows: In the formula, Represents the judgment matrix With weight vector The first new vector obtained by multiplication One component; Represents the weight vector The One component; Indicates the order of the judgment matrix; Calculate the consistency index The calculation formula is as follows: ； According to the judgment matrix order Sure The possible values ​​of ; Calculate the consistency ratio The calculation formula is as follows: In the formula, As a consistency indicator, The average random consistency index; when At that time, it is assumed that the judgment matrix has consistency. That is, the weight; otherwise, modify the judgment matrix until the consistency check is passed.

10. The method according to claim 6, characterized in that, In step four, the total human-computer interaction degree for: In the formula, and These represent the number of human-to-computer interactions and the number of computer-to-human interactions, respectively. These are adjustable parameters.

Images