A test question generation method and system based on category skill dimensions, an electronic device, and a storage medium

Abstract

The application discloses a kind of based on category skill dimension test question generation method, comprising: the subjective weight and objective weight of multiple skill dimensions are calculated;According to subjective weight and objective weight, the comprehensive weight of multiple skill dimensions is determined;The matching priority of scene template test question is calculated;Filtering the initial test question of pre-set quantity, obtain the initial test question set of multiple skill dimensions;Coverage verification step;According to test question coverage and matching priority, initial test question set is supplemented;According to the test question similarity in initial test question set after supplement, test question is deduplicated, and target test question set is obtained.The application is matched with the test question of the examination of the staff in the multiple dimensions skill comprehensive static dimension and dynamic dimension, so as to improve the comprehensiveness and scientificity of evaluation result, so that skill evaluation is more stereoscopic, objective, and can more truly reflect the actual work efficiency of service personnel.The application also discloses a system, electronic equipment and storage medium for realizing the above method.

Description

Technical Field

[0001] This invention relates to the field of computer data processing technology, and in particular to a method, system, electronic device, and storage medium for generating test questions based on category skills. Background Technology

[0002] With the rapid development of internet technology and the rise of the sharing economy, various skill-based service platforms (such as O2O platforms for on-site repair, housekeeping, and home installation) have become increasingly popular, greatly facilitating people's daily lives. These platforms typically gather a large number of individual service providers (hereinafter referred to as "service providers"). Users place orders through the platform, and service providers accept the orders and provide the corresponding offline services. To ensure service quality, protect user rights, and enhance the platform's reputation, conducting a scientific and objective evaluation of the skill levels of the service providers is one of the key aspects of platform operation.

[0003] Evaluation indicators can reflect a technician's basic knowledge and skills to a certain extent, playing a crucial role in ensuring service compliance. However, existing evaluations often focus on assessing a technician's mastery of theoretical knowledge, established operating procedures, solutions to known risks, and the use of common tools. These indicators are mostly reflections of past or present capabilities, summative assessments of existing experience, and rely excessively on static indicators while relatively neglecting the assessment of dynamic capabilities, resulting in insufficient comprehensiveness and scientific rigor in the evaluation results. Summary of the Invention

[0004] To address the aforementioned problems in the existing technology, this invention provides a method, system, electronic device, and storage medium for generating test questions based on category-based skills. The technical problem to be solved by this invention is achieved through the following technical solution:

[0005] The first aspect of this invention provides a method for generating test questions based on category skills, comprising the following steps:

[0006] The subjective and objective weights of multiple skill dimensions are determined based on the judgment matrix of the current category, the risk level in the attribute feature vector, the scenario complexity, and the category historical operation skill dimension information entropy for each skill dimension of the current category; wherein, the elements in the judgment matrix represent the importance of one skill dimension of the current category relative to another skill dimension.

[0007] The combined weights for multiple skill dimensions are determined based on the subjective and objective weights.

[0008] The matching priority of the scenario template questions is determined based on the attribute feature vector of the current product category and the comprehensive weight.

[0009] Based on the comprehensive weight, the total number of target test questions, and the matching priority, a preset number of initial test questions are selected from the scenario template test questions to obtain an initial test question set for multiple skill dimensions;

[0010] Coverage verification steps: Determine the item coverage based on the total number of target items and the number of items in the initial item set;

[0011] The initial question set is supplemented based on the question coverage and the matching priority.

[0012] Based on the question similarity in the supplemented initial question set, question deduplication is performed, and the coverage verification step is returned until the question coverage meets the first preset condition and the question similarity meets the second preset condition, thus obtaining the target question set;

[0013] The adjustment value of the comprehensive weight is determined based on the test results of the target test question set and the technology update rate in the attribute feature vector of the current product category;

[0014] The difference between the adjusted value and the overall weight is calculated to obtain the adjusted overall weight.

[0015] In one embodiment of the present invention, determining the subjective and objective weights of multiple skill dimensions based on the judgment matrix of the current category, the risk level in the attribute feature vector, the scenario complexity, and the category historical operation skill dimension information entropy for each skill dimension of the current category includes:

[0016] Calculate the maximum eigenvalue and corresponding eigenvector of the judgment matrix for the current category, and then normalize them to obtain the initial subjective weights for multiple skill dimensions;

[0017] The initial subjective weights are corrected based on the risk level and scenario complexity in the attribute feature vector of the current product category to obtain the corrected subjective weights;

[0018] The category historical task sample matrix for the current category is used to determine the category historical task skill dimension information entropy for each skill dimension; wherein, the elements in the category historical task sample matrix represent the performance score of each historical task sample in each skill dimension;

[0019] The objective weights of multiple skill dimensions for the current category are determined based on the information entropy of the historical operation skill dimension of the category.

[0020] In one embodiment of the present invention, determining the matching priority of the scenario template test questions based on the attribute feature vector of the current product category and the comprehensive weight includes:

[0021] The difficulty level of the test questions and the scene adaptability of each scene for each skill dimension are determined based on the standardization, tool dependence and scene complexity in the attribute feature vector of the current category.

[0022] The matching priority of scenario template questions is determined based on the comprehensive weight, the difficulty level of the questions, and the scenario suitability.

[0023] In one embodiment of the present invention, determining the difficulty level of the test questions and the scenario adaptability of each scenario for each skill dimension based on the standardization degree, tool dependence, and scenario complexity in the attribute feature vector of the current product category includes:

[0024] The difficulty level of the test questions is determined based on the degree of standardization and tool dependence in the attribute feature vectors of the current product category.

[0025] The scene adaptability of each skill dimension for each scene is determined based on the scene complexity.

[0026] In one embodiment of the present invention, supplementing the initial question set according to the question coverage and the matching priority includes:

[0027] When the test question coverage is less than a preset threshold, the initial test question set is supplemented according to the matching priority.

[0028] In one embodiment of the present invention, the formula for calculating the corrected subjective weight is:

[0029]

[0030] in, Indicates the first j Subjective weighting for each skill dimension. Indicates the first j Initial subjective weights for each skill dimension, Indicates the risk level. Indicates the complexity of the scene. This represents the correction factor.

[0031] In one embodiment of the present invention, the formula for calculating the comprehensive weight is:

[0032]

[0033] in, Indicates the first j The overall weight of each skill dimension, Represents the balance coefficient. Indicates the first j Objective weighting of each skill dimension.

[0034] A second aspect of this invention provides a test question generation system based on category skill dimensions, comprising:

[0035] The first determining module is used to determine the subjective and objective weights of multiple skill dimensions based on the judgment matrix of the current category, the risk level in the attribute feature vector, the scenario complexity, and the category historical operation skill dimension information entropy of each skill dimension of the current category; wherein, the elements in the judgment matrix represent the importance of one skill dimension of the current category relative to another skill dimension.

[0036] The second determining module is used to determine the comprehensive weight of multiple skill dimensions based on the subjective weight and the objective weight;

[0037] The third determining module is used to determine the matching priority of the scenario template test questions based on the attribute feature vector of the current product category and the comprehensive weight.

[0038] The filtering module is used to filter a preset number of initial questions from the scenario template questions according to the comprehensive weight, the total number of target questions and the matching priority, so as to obtain an initial question set with multiple skill dimensions.

[0039] The verification module is used for the coverage verification step: determining the question coverage based on the total number of target questions and the number of questions in the initial question set;

[0040] The supplementary module is used to supplement the initial question set according to the question coverage and the matching priority;

[0041] The deduplication module is used to deduplicat test questions based on the similarity of test questions in the supplemented initial test question set, and return to the coverage verification step until the test question coverage meets the first preset condition and the test question similarity meets the second preset condition, so as to obtain the target test question set.

[0042] The fourth determining module is used to determine the adjustment value of the comprehensive weight based on the test results of the target test question set and the technology update rate in the attribute feature vector of the current category;

[0043] The adjustment module is used to calculate the difference between the adjustment value and the comprehensive weight to obtain the adjusted comprehensive weight.

[0044] A third aspect of the present invention provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, it implements a test question generation method based on a category skill dimension provided in the first aspect of the present invention.

[0045] A fourth aspect of the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements a test question generation method based on a category skill dimension provided in the first aspect of the present invention.

[0046] The beneficial effects of this invention are:

[0047] This invention uses a multi-dimensional approach, combining static and dynamic skills to match test questions for assessors. It extends the assessment perspective to flexibility, efficiency, results-orientedness, and growth potential in actual service processes. This transforms the assessment result from a simple "compliance" judgment into a comprehensive portrait of the service personnel's overall capabilities, effectively compensating for the shortcomings of traditional assessment methods. By combining the assessor's skill level and category characteristics, personalized matching is achieved through priority calculation. Simultaneously, a similarity deduplication and differentiation update mechanism is introduced to ensure the differentiation and effectiveness of the test questions, thereby improving the comprehensiveness and scientific rigor of the assessment results. This makes skills assessment more three-dimensional, objective, and truly reflects the actual work efficiency of service personnel.

[0048] Other features and advantages of the invention will be set forth in the description which follows, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention may be realized and obtained by means of the structures particularly pointed out in the written description, claims, and drawings.

[0049] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0050] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings:

[0051] Figure 1 A flowchart illustrating a test question generation method based on category skill dimension provided in an embodiment of the present invention;

[0052] Figure 2 This is a schematic diagram of a test question generation system based on category skill dimension provided in an embodiment of the present invention. Detailed Implementation

[0053] The present invention will be further described in detail below with reference to specific embodiments, but the implementation of the present invention is not limited thereto.

[0054] like Figure 1 As shown, the first aspect of this invention provides a method for generating test questions based on category skills, comprising the following steps:

[0055] Step 11: Determine the subjective and objective weights of multiple skill dimensions based on the judgment matrix of the current category, the risk level in the attribute feature vector, the scenario complexity, and the information entropy of the category's historical operation skill dimension for each skill dimension of the current category.

[0056] In this matrix, the elements represent the importance of one skill dimension relative to another skill dimension for the current product category.

[0057] Step 12: Determine the comprehensive weight of multiple skill dimensions based on subjective and objective weights.

[0058] Step 13: Determine the matching priority of the scenario template questions based on the attribute feature vector and comprehensive weight of the current product category.

[0059] Step 14: Based on the comprehensive weight, the total number of target test questions, and the matching priority, select a preset number of initial test questions from the scenario template test questions to obtain a set of initial test questions with multiple dimensions.

[0060] Step 15, Coverage Verification Step: Determine the item coverage based on the total number of target items and the number of items in the initial item set.

[0061] Step 16: Supplement the initial question set according to question coverage and matching priority.

[0062] Step 17: Based on the question similarity in the supplemented initial question set, perform question deduplication and return to the coverage verification step until the question coverage meets the first preset condition and the question similarity meets the second preset condition, thus obtaining the target question set.

[0063] Step 18: Determine the adjustment value of the comprehensive weight based on the test results of the target test question set and the technology update rate in the attribute feature vector of the current category.

[0064] Step 19: Calculate the difference between the adjusted value and the overall weight to obtain the adjusted overall weight.

[0065] In this embodiment, the assessment of onboard personnel is conducted by matching test questions based on multiple dimensions of skills, both static and dynamic. This extends the evaluation perspective to the flexibility, efficiency, results-oriented approach, and growth potential in the actual service process. This transforms the assessment result from a simple "compliance" judgment into a comprehensive portrait of the assessor's overall capabilities, effectively compensating for the shortcomings of traditional assessment methods. By combining the assessor's skill level and category characteristics, personalized matching is achieved through priority calculation. Simultaneously, a similarity deduplication and differentiation update mechanism is introduced to ensure the differentiation and effectiveness of the test questions. This improves the comprehensiveness and scientific rigor of the assessment results, making skills assessment more three-dimensional, objective, and more accurately reflecting the actual work performance of the assessor.

[0066] Based on the first aspect of the present invention, the second aspect of the present invention provides a more detailed description of a test question generation method based on category skill dimensions. The second aspect of the present invention provides a test question generation method based on category skill dimensions, comprising the following steps:

[0067] Step 21: Determine the subjective and objective weights of multiple skill dimensions based on the judgment matrix of the current category, the risk level in the attribute feature vector, the scenario complexity, and the information entropy of the category's historical operation skill dimension for each skill dimension of the current category.

[0068] Specifically, step 21 includes steps 211-214:

[0069] Step 211: Calculate the maximum eigenvalue and corresponding eigenvector of the judgment matrix for the current category, and then normalize them to obtain the initial subjective weights for multiple skill dimensions.

[0070] The judgment matrix is ​​a multidimensional matrix. Each element in the judgment matrix represents the importance of one skill dimension of the current category relative to another skill dimension. The values ​​of the elements adopt the 1-9 scale, and each value is a specific numerical value determined by expert experience.

[0071] Here, each category has multiple skill dimensions, such as electrical products and housekeeping products. For example, each category has eight skill dimensions: theoretical understanding, standardized operation, risk management, tool usage, scenario adaptability, efficiency optimization, quality control, and continuous improvement. Therefore, the judgment matrix is ​​an 8×8 matrix. , Indicates the first i The skill dimension is relative to the first j The importance of each skill dimension.

[0072] In this step, the Analytic Hierarchy Process (AHP) algorithm is used to calculate the initial subjective weights: the maximum eigenvalue of the judgment matrix for a given category is calculated. And the corresponding feature vectors, after normalization, yield the initial subjective weights for a given product category. , n Let be a natural number representing the number of skill dimensions. Here, we'll take eight skill dimensions as an example. .

[0073] Here are the details for the eight skill dimensions:

[0074] Theoretical understanding: The degree of mastery of the core principles of the product category, industry standards, and basic concepts (such as Ohm's law for electrical circuits and the principles of disinfection in housekeeping).

[0075] Standardized operations: standardized operating procedures, compliance of operating steps, and quality acceptance standards (such as wiring techniques for electricians and clothing sorting and washing procedures for housekeepers).

[0076] Risk response: prediction of emergencies, troubleshooting, and emergency response capabilities (such as electrician's first aid for electric shock and housekeeping's emergency response to gas leaks).

[0077] Tool usage: Selection, operation, maintenance and calibration of professional tools / equipment (such as the use of multimeters by electricians and the maintenance of vacuum cleaners by housekeepers).

[0078] Scenario adaptability: The flexibility and adaptability of adjusting operation plans based on different application scenarios (such as environment, user needs, constraints) (e.g., the wiring differences for electricians in humid vs. dry environments, and the adjustment of disinfection plans for housekeeping in households with infants).

[0079] Efficiency optimization capability: The ability to improve operational efficiency through process optimization and resource allocation while ensuring compliance (such as multi-circuit parallel construction schemes for electricians and whole-house cleaning path planning for housekeeping services).

[0080] Quality control capabilities: Self-inspection / mutual inspection during the work process, error control, and the ability to implement the results acceptance standards (such as electrical wiring insulation resistance testing and housekeeping inspection of cleaning dead corners).

[0081] Continuous improvement capability: The ability to iterate operating methods and supplement knowledge reserves based on work feedback and new technologies / standards (such as electricians learning wiring specifications for new intelligent devices and housekeepers learning how to use environmentally friendly cleaning agents).

[0082] Step 212: Correct the initial subjective weights based on the risk level and scenario complexity in the attribute feature vector of the current product category to obtain the corrected subjective weights.

[0083] In this step, the attribute feature vectors of the product category are pre-calculated. The specific calculation process for the attribute feature vectors is as follows: Define the five core attribute vectors of the product category. X =[ ], respectively:

[0084]

[0085]

[0086]

[0087]

[0088]

[0089] right X=[ Feature normalization is performed to obtain attribute feature vectors. The normalization calculation formula is as follows:

[0090] =

[0091] in, Core attribute vector X The minimum value, Core attribute vector X The maximum value.

[0092] Combining normalization features (Risk Level) (Scenario Complexity) The initial subjective weights are adjusted, and the weights of the "Risk Response and Scenario Adaptation" dimensions for high-risk and high-complexity categories are increased. The formula for calculating the adjusted subjective weights for a given category is as follows:

[0093]

[0094] in, Indicates the first j Subjective weighting after adjusting each skill dimension , To correct the coefficients, normalize again after correction to ensure... .

[0095] Step 213: Determine the category historical task skill dimension information entropy for each skill dimension based on the category historical task sample matrix of the current category.

[0096] Among them, the elements in the category historical task sample matrix represent the performance score of each historical task sample in each skill dimension.

[0097] In this step, the first step is to construct a category-specific historical operation sample matrix. ,in m This represents the number of historical operational samples for this product category. For the first k The sample at the th j Performance scores for each skill dimension, ( (Values ​​range from 0 to 100). Then calculate the... j Skill-based category history of assignments, skill-based information entropy ,in (like =0, then =0).

[0098] Step 214: Determine the objective weights of multiple skill dimensions for the current category based on the information entropy of the historical operation skill dimension of the category.

[0099] In this step, the first j The formula for calculating the objective weight of the skill dimension is as follows:

[0100]

[0101] Indicates the first j Objective weights for the skill dimension. In this embodiment, the entropy weight method is used to calculate the objective weights.

[0102] Step 22: Determine the comprehensive weight of multiple skill dimensions based on subjective and objective weights.

[0103] In this step, a linear weighting method is used to integrate subjective and objective weights, taking into account both expert experience and data objectivity:

[0104]

[0105] in, Indicates the first j The overall weighting of the skills dimension, α The balance coefficient is set at 0.5 ≤ α ≤ 0.7, adjusted according to category characteristics. For example, for high-risk categories, α = 0.7, emphasizing expert experience; for general categories, α = 0.5, emphasizing data patterns. The coefficient must also satisfy the following conditions: .

[0106] For example, the calculated combined weights for electrical products and household goods are shown in Table 1:

[0107] Table 1

[0108]

[0109] Step 23: Determine the matching priority of the scenario template questions based on the attribute feature vector and comprehensive weight of the current product category.

[0110] Step 23 includes steps 231-233:

[0111] Step 231: Determine the difficulty level of the test questions based on the degree of standardization and tool dependence in the attribute feature vector of the current product category.

[0112] In this step, a scenario-based test question template library is built in advance. , d Difficulty levels (1 = Basic, 2 = Intermediate, 3 = Advanced). sBased on scenario types, one skill dimension can correspond to multiple scenario types. For example, the "Risk Response" skill dimension in the electrical engineering category includes scenarios such as electric shock, short circuit, and overload. Each scenario corresponds to a scenario-based test question template. Each test question template... It includes four core elements: question description, assessment points, scoring criteria, and reference analysis.

[0113] Combination (Degree of standardization) and (Tool Dependency) Calculate the difficulty level of test questions The calculation formula is:

[0114]

[0115] in, To round up, the results are mapped to levels 1-3. Therefore, categories with high standardization and high tool dependence (such as electricians) require more difficult test questions; categories with low standardization (such as housekeeping) have moderate difficulty.

[0116] Step 232: Determine the scene adaptability of each scene for each skill dimension based on scene complexity.

[0117] In this step, combined (Scenario complexity), calculate the fit of the question template for each scenario:

[0118]

[0119] in, (Scene Complexity Coefficient) is assigned a value (0.3-0.9) based on the complexity of the scene. Each skill dimension includes multiple scenes, and each scene has a complexity coefficient to ensure that categories with high scene complexity are matched with more complex scene test templates.

[0120] Step 233: Determine the matching priority of scenario template questions based on comprehensive weight, question difficulty level, and scenario adaptability.

[0121] In this step, the matching priority of each dimension of the scene is determined. The calculation formula is as follows:

[0122]

[0123] in, The difficulty matching score (0-1) is determined by the degree of fit between the technical difficulty of the product category and the difficulty of the template. It is a pre-set value, for example, , , Each scenario calculates a corresponding matching priority.

[0124] Step 24: Based on the comprehensive weight, the total number of target test questions and the matching priority, select a preset number of scenario template test questions for each skill dimension as initial test questions to obtain an initial test question set for multiple skill dimensions.

[0125] In this step, the total number of target test questions for a certain product category is preset. The required number of test questions for each skill dimension is calculated by multiplying the comprehensive weight of each skill dimension by the total number of target test questions. The matching priority of each skill dimension is sorted in descending order, and the scenario template test questions with the preset number of names before sorting are selected as the initial test questions. After multiple skill dimensions are selected, the initial test question set is obtained. For example, if the total number of target test questions for the electrical products category is 20, and the overall weight of the tool usage dimension is 0.15, then the tool usage dimension needs to match 3 questions in the template library. There are 5 scenarios for the tool usage dimension, corresponding to 5 template test questions. Then, 5 matching priorities are calculated, and the scenario template test questions of the top 3 scenarios in the matching priority ranking are selected.

[0126] Step 25, Coverage Verification Step: Determine the item coverage based on the total number of target items and the number of items in the initial item set.

[0127] In this step, the formula for calculating the test item coverage is:

[0128]

[0129] in, K The target total number of test questions for the current product category. To round up, This indicates the number of items in the initial test set.

[0130] Step 26: Supplement the initial question set according to question coverage and matching priority.

[0131] In this step, specifically, when the question coverage is less than a preset threshold, the initial question set is supplemented according to the matching priority. When necessary, scenario template questions for skill dimensions, ranked from highest to lowest overall weight, are added to the test. Within each skill dimension, additions are made based on matching priority. For example, if additional questions are needed for the electrical engineering category, the risk response skill dimension has the highest overall weight. Therefore, the question with the fourth highest matching priority in the corresponding scenario questions for the risk response skill dimension is selected (the top three matching priority questions are already in the initial question set; the next highest matching priority question is selected if it is not in the initial question set). If there are only three risk response questions, and they have already been selected as initial questions, then the next highest matching priority question is selected from the skill dimensions with the next highest overall weight, and so on.

[0132] Step 27: Based on the question similarity in the supplemented initial question set, perform question deduplication, return to step 26, until the question coverage meets the first preset condition and the question similarity meets the second preset condition, and obtain the target question set.

[0133] After adding a question in step 26, deduplication is performed, removing questions with lower matching priority to ensure question differentiation. The question similarity in the initial question set after addition is then calculated. The calculation formula is:

[0134]

[0135] in, This indicates any two different questions following the current supplementary question. Indicates a statistical quantity.

[0136] Based on the content of the questions, a certain number of keywords can be identified for each question. The similarity is calculated using the aforementioned method for the number of keywords corresponding to two questions, which is the percentage of identical keywords in the two questions out of the total number of keywords in both questions. Questions with a similarity greater than 0.7 and lower matching priority are deleted. After deduplication, since the number of questions is reduced, the process returns to step 26 to supplement the questions. Deduplication is then performed again after supplementation until the question coverage meets the first preset condition and the question similarity meets the second preset condition. The supplementation and deduplication process is then complete, yielding the target question set. In this example, the first preset condition is a coverage greater than or equal to 0.9, and the second preset condition is a similarity less than or equal to 0.7.

[0137] After coverage verification and deduplication, the final target question set is output and sorted according to the comprehensive weight of the skill dimension (questions of the skill dimension with higher weight are ranked higher).

[0138] Step 28: Determine the adjustment value of the comprehensive weight based on the test results of the target test question set and the technology update rate in the attribute feature vector of the current product category.

[0139] In this step, the adjustment value of the overall weight. The calculation formula is:

[0140]

[0141] in, For the average score, , Indicates the first j Test results in the skills dimension, This indicates the rate of technological updates.

[0142] Here, product categories with faster technological updates have larger weight iterations, adapting to technological changes.

[0143] Step 29: Calculate the difference between the adjusted value and the overall weight to obtain the adjusted overall weight.

[0144] In this step, the adjusted value of the overall weight for a skill dimension is less than 0 and... When the absolute value is greater than 0.1, it indicates that the score for that skill dimension is significantly lower than the average level and the gap between the score and the average score is large, suggesting that this dimension needs to be strengthened. In this case, the difference between the adjustment value and the overall weight is calculated, and the overall weight of that dimension is increased. Conversely, the overall weight is decreased. The adjustment value of the overall weight of a skill dimension is greater than or equal to 0 and... When the absolute value is greater than 0.1, it indicates that the score for this skill dimension is significantly higher than the average level and is relatively high. This suggests that the assessment of this dimension can be weakened. Therefore, the difference between the adjustment value and the overall weight is calculated, and the overall weight of this dimension is reduced. After adjusting the overall weight, return to step 22 to rematch the level and scenario test templates for periodic re-examination.

[0145] Here, when adjusting the overall weights, if some of the overall weights are increased or decreased, in order to ensure... The minimum value of the adjusted comprehensive weight is further adjusted based on the difference between the sum of all adjusted comprehensive weights and 1, to ensure... .

[0146] Step 30: Calculate the scores of resident personnel with different skill levels for each scenario test question, obtain the discrimination index of each scenario test question, and update the content of the scenario test questions based on the discrimination index, which is to update the scenario test question template library.

[0147] In this step, the discrimination factor The calculation formula is:

[0148]

[0149] in, The average score of high-skilled test takers on scenario-based test item t. Let S be the average score of those with low skill levels, and S be the standard deviation of the total scores of all items for those with high and low skill levels. If the score is less than 0.3 (low discrimination), update the content or scenario of the scenario-based test questions; if... A score >0.7 (high discrimination) is considered for inclusion in the core template library and prioritized for matching. High-skill level applicants are those who score over 80 points in the pre-test, while low-skill level applicants are those who score below 50 points. The pre-test is an initial self-assessment required by applicants during registration, conducted before the target test questions in this embodiment.

[0150] In this embodiment, four dynamic dimensions—"scenario adaptation, efficiency optimization, quality control, and continuous improvement"—are added to the multidimensional skills assessment system. This addresses the industry pain point of traditional assessments that "emphasize static indicators while neglecting dynamic capabilities," and upgrades skills assessment from "compliance" to "comprehensiveness," covering the entire lifecycle needs of professional skills.

[0151] The combined weighting method of AHP and entropy weighting incorporates experts' experience in assessing the core risks of the category and objectively adjusts the weights based on historical data, avoiding the one-sidedness of single weighting. At the same time, through the dimensional weight iteration mechanism, the weights are dynamically adjusted according to the evaluation feedback, breaking through the limitations of the traditional "fixed weight".

[0152] A three-dimensional template library of "dimension-difficulty-scenario" is constructed. Combining the skill level of the person being assessed with the characteristics of the category, personalized matching is achieved through priority calculation, which solves the problem of the "one-size-fits-all" approach of traditional template libraries. At the same time, a similarity deduplication and differentiation update mechanism is introduced to ensure the differentiation and effectiveness of the test questions.

[0153] The algorithm is not statically decomposed, but continuously optimizes the dimension weights and test question template library through evaluation feedback, realizing a virtuous cycle of "evaluation-improvement". It adapts to the technological updates and scenario changes of different categories (such as the popularization of smart devices in the electrical industry and the upgrading of environmental protection standards in the housekeeping industry), and has strong scalability and adaptability.

[0154] like Figure 2 As shown, a third aspect of the present invention provides a test question generation system based on category skill dimensions, comprising:

[0155] The first determining module 41 is used to determine the subjective and objective weights of multiple skill dimensions based on the judgment matrix of the current category, the risk level in the attribute feature vector, the scenario complexity, and the category historical operation skill dimension information entropy of each skill dimension of the current category; wherein, the elements in the judgment matrix represent the importance of one skill dimension of the current category relative to another skill dimension.

[0156] The second determining module 42 is used to determine the comprehensive weight of multiple skill dimensions based on subjective weight and objective weight;

[0157] The third determining module 43 is used to determine the matching priority of scenario template questions based on the attribute feature vector and comprehensive weight of the current product category.

[0158] The filtering module 44 is used to filter a preset number of initial questions from the scenario template questions based on the comprehensive weight, the total number of target questions and the matching priority, so as to obtain an initial question set with multiple skill dimensions.

[0159] Verification module 45 is used for the coverage verification step: determining the question coverage based on the total number of target questions and the number of questions in the initial question set;

[0160] Supplement module 46 is used to supplement the initial question set based on question coverage and matching priority;

[0161] The deduplication module 47 is used to deduplicat test questions based on the similarity of test questions in the supplemented initial test question set, and return to the coverage verification step until the test question coverage meets the first preset condition and the test question similarity meets the second preset condition, so as to obtain the target test question set.

[0162] The fourth determining module 48 is used to determine the adjustment value of the comprehensive weight based on the test results of the target test question set and the technology update rate in the attribute feature vector of the current category;

[0163] Adjustment module 49 is used to calculate the difference between the adjustment value and the overall weight to obtain the adjusted overall weight.

[0164] In one embodiment of the present invention, subjective and objective weights for multiple skill dimensions are determined based on the judgment matrix of the current category, the risk level in the attribute feature vector, the scenario complexity, and the category historical operation skill dimension information entropy for each skill dimension of the current category, including:

[0165] Calculate the maximum eigenvalue and corresponding eigenvector of the judgment matrix for the current category, and then normalize them to obtain the initial subjective weights for multiple skill dimensions;

[0166] The initial subjective weights are corrected based on the risk level and scenario complexity in the attribute feature vector of the current product category to obtain the corrected subjective weights.

[0167] The category historical task sample matrix is ​​used to determine the category historical task skill dimension information entropy for each skill dimension; where the elements in the category historical task sample matrix represent the performance score of each historical task sample in each skill dimension.

[0168] The objective weights of multiple skill dimensions for the current product category are determined based on the information entropy of historical operational skill dimensions for that category.

[0169] In one embodiment of the present invention, determining the matching priority of scenario template questions based on the attribute feature vector and comprehensive weight of the current product category includes:

[0170] The difficulty level of the test questions and the scene adaptability of each scene for each skill dimension are determined based on the standardization, tool dependence and scene complexity in the attribute feature vector of the current category.

[0171] The matching priority of scenario template questions is determined based on comprehensive weight, question difficulty level, and scenario adaptability.

[0172] In one embodiment of the present invention, determining the difficulty level of test questions and the scenario adaptability of each scenario for each skill dimension based on the standardization degree, tool dependence, and scenario complexity in the attribute feature vector of the current product category includes:

[0173] The difficulty level of the test questions is determined based on the degree of standardization and tool dependence in the attribute feature vectors of the current product category.

[0174] The scene adaptability of each skill dimension for each scene is determined based on the scene complexity.

[0175] In one embodiment of the present invention, the initial question set is supplemented according to question coverage and matching priority, including:

[0176] When the coverage of test questions is less than a preset threshold, the initial test question set is supplemented according to the matching priority.

[0177] In one embodiment of the present invention, the formula for calculating the corrected subjective weight is as follows:

[0178]

[0179] in, Indicates the first j Subjective weighting for each skill dimension. Indicates the first j Initial subjective weights for each skill dimension, Indicates the risk level. Indicates the complexity of the scene. This represents the correction factor.

[0180] In one embodiment of the present invention, the formula for calculating the comprehensive weight is as follows:

[0181]

[0182] in, Indicates the first j The overall weight of each skill dimension, Represents the balance coefficient. Indicates the first j Objective weighting of each skill dimension.

[0183] A fourth aspect of the present invention provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, it implements the above-described method for generating test questions based on a category skill dimension provided by the present invention.

[0184] A fifth aspect of the present invention also provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of the test question generation method based on category skill dimension provided in the above-described embodiments of the present invention.

[0185] The memory may include random access memory (RAM) or non-volatile memory (NVM), such as at least one disk storage device. Optionally, the memory may also be at least one storage system located remotely from the aforementioned processor.

[0186] The processors mentioned above can be general-purpose processors, including central processing units (CPUs), network processors (NPs), etc.; they can also be digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware systems.

[0187] The method provided in this invention can be applied to electronic devices. Specifically, the electronic device can be a desktop computer, a portable computer, a smart mobile terminal, a server, etc. No limitation is made herein; any electronic device that can implement this invention falls within the protection scope of this invention.

[0188] For system / electronic device embodiments, since they are basically similar to method embodiments, the description is relatively simple, and relevant parts can be found in the description of the method embodiments.

[0189] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A system that specifies functions in one or more boxes.

[0190] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0191] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0192] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A method for generating test questions based on category skills, characterized in that, Includes the following steps: The subjective and objective weights of multiple skill dimensions are determined based on the judgment matrix of the current category, the risk level in the attribute feature vector, the scenario complexity, and the category historical operation skill dimension information entropy for each skill dimension of the current category; wherein, the elements in the judgment matrix represent the importance of one skill dimension of the current category relative to another skill dimension. The combined weights for multiple skill dimensions are determined based on the subjective and objective weights. The matching priority of the scenario template test questions is determined based on the attribute feature vector of the current product category and the comprehensive weight. Based on the comprehensive weight, the total number of target test questions, and the matching priority, a preset number of initial test questions are selected from the scenario template test questions to obtain an initial test question set for multiple skill dimensions; Coverage verification steps: Determine the item coverage based on the total number of target items and the number of items in the initial item set; The initial question set is supplemented based on the question coverage and the matching priority. Based on the question similarity in the supplemented initial question set, question deduplication is performed, and the coverage verification step is returned until the question coverage meets the first preset condition and the question similarity meets the second preset condition, thus obtaining the target question set; The adjustment value of the comprehensive weight is determined based on the test results of the target test question set and the technology update rate in the attribute feature vector of the current product category; Calculate the difference between the adjusted value and the overall weight to obtain the adjusted overall weight; The step of determining the matching priority of scenario template questions based on the attribute feature vector of the current product category and the comprehensive weight includes: The difficulty level of the test items is determined based on the degree of standardization and tool dependence in the attribute feature vector of the current product category; the difficulty level of the test items The calculation formula is: in, To round up, This indicates the degree of standardization in the attribute feature vector of the current product category. This indicates the degree of standardization and tool dependence in the attribute feature vector of the current product category; The scenario adaptability of each scenario is determined based on the scenario complexity; the scenario adaptability of each scenario... The calculation formula is: in, Indicates the complexity of the scene. Indicates the first j Skill dimension s The scene complexity coefficient of the type of scenario; The matching priority of scenario template questions is determined based on the comprehensive weight, the difficulty level of the questions, and the scenario suitability; j Skill dimension s Matching priority of scenario template questions for the specified scenario type The calculation formula is as follows: in, Indicates the first j The overall weighting of the skills dimension, Indicates the first j Difficulty level of questions in the skills dimension The difficulty level of the match.

2. The method as described in claim 1, characterized in that, The process of determining the subjective and objective weights of multiple skill dimensions based on the judgment matrix of the current category, the risk level in the attribute feature vector, the scenario complexity, and the information entropy of the category's historical operation skill dimension for each skill dimension of the current category includes: Calculate the maximum eigenvalue and corresponding eigenvector of the judgment matrix for the current category, and then normalize them to obtain the initial subjective weights for multiple skill dimensions; The initial subjective weights are corrected based on the risk level and scenario complexity in the attribute feature vector of the current product category to obtain the corrected subjective weights; The category historical task sample matrix for the current category is used to determine the category historical task skill dimension information entropy for each skill dimension; wherein, the elements in the category historical task sample matrix represent the performance score of each historical task sample in each skill dimension; The objective weights of multiple skill dimensions for the current category are determined based on the information entropy of the historical operation skill dimension of the category.

3. The method as described in claim 2, characterized in that, The process of determining the difficulty level of test questions and the scenario adaptability of each scenario for each skill dimension based on the standardization degree, tool dependence, and scenario complexity in the attribute feature vector of the current product category includes: The difficulty level of the test questions is determined based on the degree of standardization and tool dependence in the attribute feature vectors of the current product category. The scene adaptability of each skill dimension for each scene is determined based on the scene complexity.

4. The method as described in claim 1, characterized in that, The process of supplementing the initial question set according to the question coverage and the matching priority includes: When the test question coverage is less than a preset threshold, the initial test question set is supplemented according to the matching priority.

5. The method as described in claim 2, characterized in that, The formula for calculating the corrected subjective weight is as follows: in, Indicates the first j Subjective weighting for each skill dimension. Indicates the first j Initial subjective weights for each skill dimension, Indicates the risk level. Indicates the complexity of the scene. This represents the correction factor.

6. The method as described in claim 5, characterized in that, The formula for calculating the overall weight is as follows: in, Indicates the first j The overall weight of each skill dimension, Represents the balance coefficient. Indicates the first j Objective weighting of each skill dimension.

7. A test question generation system based on category skill dimension, characterized in that, include: The first determining module is used to determine the subjective and objective weights of multiple skill dimensions based on the judgment matrix of the current category, the risk level in the attribute feature vector, the scenario complexity, and the category historical operation skill dimension information entropy of each skill dimension of the current category; wherein, the elements in the judgment matrix represent the importance of one skill dimension of the current category relative to another skill dimension. The second determining module is used to determine the comprehensive weight of multiple skill dimensions based on the subjective weight and the objective weight; The third determining module is used to determine the matching priority of the scenario template test questions based on the attribute feature vector of the current product category and the comprehensive weight. The filtering module is used to filter a preset number of initial questions from the scenario template questions according to the comprehensive weight, the total number of target questions and the matching priority, so as to obtain an initial question set with multiple skill dimensions. The verification module is used for the coverage verification step: determining the question coverage based on the total number of target questions and the number of questions in the initial question set; The supplementary module is used to supplement the initial question set according to the question coverage and the matching priority; The deduplication module is used to deduplicat test questions based on the similarity of test questions in the supplemented initial test question set, and return to the coverage verification step until the test question coverage meets the first preset condition and the test question similarity meets the second preset condition, so as to obtain the target test question set. The fourth determining module is used to determine the adjustment value of the comprehensive weight based on the test results of the target test question set and the technology update rate in the attribute feature vector of the current category; An adjustment module is used to calculate the difference between the adjustment value and the comprehensive weight to obtain the adjusted comprehensive weight; The step of determining the matching priority of scenario template questions based on the attribute feature vector of the current product category and the comprehensive weight includes: The difficulty level of the test items is determined based on the degree of standardization and tool dependence in the attribute feature vector of the current product category; the difficulty level of the test items The calculation formula is: in, To round up, This indicates the degree of standardization in the attribute feature vector of the current product category. This indicates the degree of standardization and tool dependence in the attribute feature vector of the current product category; The scenario adaptability of each scenario is determined based on the scenario complexity; the scenario adaptability of each scenario... The calculation formula is: in, Indicates the complexity of the scene. Indicates the first j Skill dimension s The scene complexity coefficient of the type of scenario; The matching priority of scenario template questions is determined based on the comprehensive weight, the difficulty level of the questions, and the scenario suitability; j Skill dimension s Matching priority of scenario template questions for the specified scenario type The calculation formula is as follows: in, Indicates the first j The overall weighting of the skills dimension, Indicates the first j Difficulty level of questions in the skills dimension The difficulty level of the match.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the test question generation method based on the category skill dimension as described in any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the test question generation method based on the category skill dimension as described in any one of claims 1 to 6.

Images