Multi-path skill assessment method and system for construction industry

Abstract

The invention discloses a multi-path skill assessment method and system for the building industry, and relates to the technical field of data processing, and the method comprises the steps: interacting a skill resource database, and extracting a multi-path skill resource data set corresponding to a target skill group, the multi-path skill resource data set comprises multi-dimensional skill resource data of multiple skills in the target skill group; a multi-path data propagation evaluation channel is constructed, and the multi-path data propagation evaluation channel comprises a plurality of groups of weak evaluation sub-channels; performing propagation evaluation on the multi-path skill resource data set based on the multi-path data propagation evaluation channel to obtain a propagation evaluation index set; extracting a concentrated value of the propagation evaluation index set, and generating a skill propagation baseline; and obtaining a skill assessment constraint of the target scene, performing grading assessment of the target skill group in combination with the skill propagation baseline, and outputting an assessment result. Therefore, the technical effects of accurately evaluating the comprehensive quality and assisting in exerting the individual advantages are achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and in particular to a multi-path skill assessment method and system for the construction industry. Background Art

[0002] With the development of construction technology and the complexity of construction projects, the requirements for the skill level of practitioners are also constantly increasing. In order to ensure the quality and safety of construction projects, it is necessary to strengthen the skill assessment and training of practitioners. However, the existing skill assessment methods mostly focus on a single skill, and the differences in the proficiency of different individuals in different skill items are not fully considered. There are technical problems such as low accuracy of comprehensive evaluation and difficulty in reflecting individual advantages. Summary of the invention

[0003] The purpose of this application is to provide a multi-path skill assessment method and system for the construction industry, so as to solve the technical problems in the prior art of low comprehensive evaluation accuracy and difficulty in reflecting individual advantages.

[0004] In view of the above technical problems, the present application provides a multi-path skill assessment method and system for the construction industry.

[0005] In a first aspect, the present application provides a multi-path skills assessment method for the construction industry, wherein the method comprises: Interactive skill resource database, extracting a multi-path skill resource data set corresponding to a target skill group, wherein the multi-path skill resource data set includes multi-dimensional skill resource data of multiple skills in the target skill group; constructing a multi-path data propagation evaluation channel, wherein the multi-path data propagation evaluation channel includes multiple groups of weak evaluation sub-channels; performing propagation evaluation of the multi-path skill resource data set based on the multi-path data propagation evaluation channel to obtain a propagation evaluation index set; extracting a central value of the propagation evaluation index set to generate a skill propagation baseline; obtaining the skill assessment constraints of the target scenario, combining the skill propagation baseline, performing a graded assessment of the target skill group, and outputting the assessment results.

[0006] In a second aspect, the present application also provides a multi-path skill assessment system for the construction industry, wherein the system comprises: A resource acquisition module, the resource acquisition module is used to interact with the skill resource database and extract a multi-path skill resource data set corresponding to the target skill group, wherein the multi-path skill resource data set includes multi-dimensional skill resource data of multiple skills in the target skill group; A propagation evaluation configuration module, wherein the propagation evaluation configuration module is used to construct a multi-path data propagation evaluation channel, wherein the multi-path data propagation evaluation channel includes a plurality of groups of weak evaluation sub-channels; A propagation evaluation module, the propagation evaluation module is used to perform a propagation evaluation of the multi-path skill resource data set based on the multi-path data propagation evaluation channel to obtain a propagation evaluation index set; A propagation base acquisition module, the propagation base acquisition module is used to extract the central value of the propagation evaluation index set to generate a skill propagation baseline; A multi-path assessment module is used to obtain the skill assessment constraints of the target scenario, combine the skill propagation baseline, conduct a graded assessment of the target skill group, and output the assessment results.

[0007] One or more technical solutions provided in this application have at least the following technical effects or advantages: By acquiring multiple network security modules connected to the cloud computing module, wherein the data of each network security module is transmitted to the cloud computing module for storage; detecting the security threat accident set of each network security module in the multiple network security modules; storing the security threat accident set in the security threat accident sharing library, which is embedded in the cloud computing module; collecting the communication protocol information between every two modules in the multiple network security modules, and performing security detection on the collected communication protocols to obtain the transmission risk index between every two modules; when the transmission risk index between the two modules is greater than the preset transmission risk index, the two modules are identified as a group, and the module identification result is output; the data in the security threat accident sharing library is shared according to the module identification result. Thus, the technical effect of accurately evaluating the comprehensive quality and helping to give play to the individual advantages is achieved.

[0008] The above description is only an overview of the technical solution of the present application. In order to more clearly illustrate the technical means of the present application and to implement it in accordance with the contents of the specification, and to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0009] The embodiments of the present invention and the following brief description are illustrated in conjunction with the accompanying drawings, which are described as follows: Figure 1 A flowchart of a multi-path skill assessment method for the construction industry used in this application; Figure 2 This is a schematic diagram of the structure of the multi-path skill assessment system used in the construction industry in this application.

[0010] Explanation of the reference numerals: resource acquisition module 11, propagation assessment configuration module 12, propagation assessment module 13, propagation base acquisition module 14, multi-path assessment module 15. DETAILED DESCRIPTION

[0011] This application solves the technical problems faced by the prior art of low comprehensive evaluation accuracy and difficulty in reflecting individual advantages by providing a multi-path skill assessment method and system for the construction industry.

[0012] The solution in the embodiment of the present technology, in order to solve the above problems, adopts the following overall ideas: First, the interactive skill resource database is used to extract the multi-path skill resource data set corresponding to the target skill group, wherein the multi-path skill resource data set includes multi-dimensional skill resource data of multiple skills in the target skill group. Next, a multi-path data propagation evaluation channel is constructed, wherein the multi-path data propagation evaluation channel includes multiple groups of weak evaluation sub-channels. Then, a propagation evaluation of the multi-path skill resource data set is performed based on the multi-path data propagation evaluation channel to obtain a propagation evaluation index set. Then, the centralized value of the propagation evaluation index set is extracted to generate a skill propagation baseline. Finally, the skill assessment constraints of the target scenario are obtained, and a graded assessment of the target skill group is performed in combination with the skill propagation baseline, and the assessment results are output.

[0013] In order to better understand the above technical solution, the above technical solution will be described in detail below in conjunction with the drawings and specific implementation methods of the specification. It should be noted that the described embodiments are only part of the embodiments of the present application, rather than all of the embodiments of the present application. It should be understood that the present application is not limited to the example embodiments described herein. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention. It should also be noted that for the sake of ease of description, only the parts related to the present invention are shown in the drawings, rather than all of them.

[0014] Embodiment 1

[0015] like Figure 1 As shown, the present application provides a multi-path skill assessment method for the construction industry, the method comprising: An interactive skill resource database is used to extract a multi-path skill resource data set corresponding to a target skill group, wherein the multi-path skill resource data set includes multi-dimensional skill resource data of multiple skills in the target skill group.

[0016] Optionally, the skill resource database refers to a database containing multiple professional skill training resources in the construction scenario, including video resources, picture resources, audio resources, text resources, model resources, etc. The target skill group refers to a specific skill combination or skill field that needs to be assessed. In other words, the target skill group includes a variety of related construction industry skills.

[0017] Optionally, an interactive skill resource database is used to obtain a multi-path skill resource data set for a target skill group, wherein the multi-path skill resource data set includes multiple groups of multi-dimensional skill resource data, each group of multi-dimensional skill resource data corresponds to a construction industry skill in the target skill group, and in addition, the multi-path skill resource data set includes multiple associated skill names, resource types and resource data.

[0018] Exemplarily, the skill resource database is a relational database, including: a skill table (Skills), which is used to store basic information of skills; a skill resource table (SkillResources), which is used to store resource data of each skill; and a resource mapping table (SkillGroupMappings): which stores the mapping relationship between skill resource data and skills.

[0019] For example, a query statement is written to extract a specific target skill group and its corresponding skills and resource data from the skill resource database. The query statement obtains the ID of the target skill group from the skill group table, then finds all the skills included in the skill group through the mapping table, and finally obtains the resource data of these skills from the skill resource table.

[0020] In some embodiments, the interactive skill resource database extracts a multi-path skill resource data set corresponding to the target skill group, including: Interact with target scenarios and acquire target skill sets.

[0021] The target skill set is analyzed in association to generate a plurality of associated skill groups.

[0022] Random extraction is performed among the multiple associated skill groups to obtain the target skill group, and the skill resource database is traversed with the target skill group as an index constraint to extract the multi-path skill resource data set.

[0023] Optionally, the target scenario refers to a target construction industry scenario that requires skill assessment, and the target skill set is a collection of all construction industry skill types that require skill assessment in the target scenario.

[0024] Optionally, a correlation analysis is performed on multiple types of construction industry skills in the target skill set to obtain multiple related skill groups, and the multiple related skill groups correspond to multiple types of work or work processes in the above target scenario. Exemplarily, an association analysis algorithm (such as cluster analysis, graph theory, association rule mining, etc.) is used to analyze the target skill set.

[0025] Specifically, the multiple types of work or work processes in the above target scenarios involve multiple construction industry skills, and each construction industry practitioner has individual differences in the proficiency of different skills. In other words, an individual practitioner with poor performance in a certain construction industry skill may have higher potential in another construction industry skill and may have better skill application performance. By obtaining multiple related skill groups, it is helpful to adapt the most suitable construction industry skills to differentiated construction industry practitioners, which helps to improve the professional ability of practitioners on the one hand, and at the same time improves the guiding value of construction industry skill assessment.

[0026] Optionally, a target skill group is randomly selected from multiple related skill groups, and the target skill group is used as an index to search for related resources in the skill resource database to ensure that the extracted resources are highly relevant to the target skill group. The obtained multi-path skill resource dataset contains multiple learning paths and resources for multiple related skills. Exemplary examples include text learning paths, video learning paths, voice learning paths, practical learning paths, etc.

[0027] A multipath data propagation evaluation channel is constructed, wherein the multipath data propagation evaluation channel includes a plurality of groups of weak evaluation sub-channels.

[0028] Optionally, the multiple groups of weak evaluation sub-channels included in the multi-path data propagation evaluation channel are respectively used to perform propagation evaluation on the resource data of multiple construction industry skills in the multi-path skill resource data set. Exemplarily, the multiple groups of weak evaluation sub-channels perform propagation evaluation through the resource data quality, resource data validity, and resource data application evaluation of the resource data in the multi-path skill resource data set, and quantitatively evaluate the learning conversion efficiency of the multi-path skill resource data set.

[0029] In some embodiments, constructing a multipath data propagation assessment channel includes: The multi-path skill resource data set is analyzed to obtain resource type distribution characteristics.

[0030] Evaluate the computing power of the target scenario and allocate computing power across multiple paths based on the distribution characteristics of the resource types.

[0031] Based on the multi-path computing power allocation result, the multi-path data propagation evaluation channel is configured.

[0032] Optionally, before analyzing the multi-path skill resource data set, first, clean the data, remove duplicate and invalid data, ensure the integrity and accuracy of the data, and then, based on the resource type tag of the multi-path skill resource data set, classify each resource into the corresponding category; then, count the number of each resource, obtain the distribution of resource types, calculate the proportion of each resource in the entire data set, and analyze the characteristics of different resource types, such as the duration of video resources, the number of words in text resources, the audio quality of voice resources, the performance of model resources, etc. Finally, integrate the above statistical analysis results and store them as resource type distribution features. The resource type distribution feature characterizes the proportional relationship between different types of resource data in the multi-path skill resource data set, which helps to allocate appropriate evaluation computing power to multiple paths for processing different types of resource data in the future.

[0033] Optionally, the computing power of the target scene is evaluated by performing a computing power evaluation on the target scene, including resources such as CPU, GPU, and memory, to obtain the evaluated computing power of the target scene.

[0034] Furthermore, according to the computing power evaluation results, the computing power of multiple paths is reasonably allocated to balance the performance of each path, so that the resource types with a larger amount of resources or a larger proportion of resources correspond to a higher computing power allocation ratio and have a higher evaluation processing speed. Through the above computing power allocation, it is possible to avoid a certain path from being overloaded or idle. This helps to synchronize the output of evaluation results and complete the evaluation of multiple paths in the subsequent evaluation, thereby improving the efficiency of propagation evaluation.

[0035] Furthermore, based on the multi-path computing power allocation results, after configuring the multi-path data propagation evaluation channel, a feedback mechanism is established to timely obtain the computing power usage status during the evaluation process as feedback results, and according to the feedback results, dynamically adjust the computing power allocation and configuration of each sub-channel to optimize the evaluation efficiency and accuracy.

[0036] In some implementations, configuring the multi-path data propagation evaluation channel based on the multi-path computing power allocation result includes: Based on the resource type distribution characteristics, multiple groups of weak evaluation sub-channels are constructed and trained, and the multiple groups of weak evaluation sub-channels correspond to multiple resource types.

[0037] A first group of weak evaluation sub-channels is extracted, and group performance consumption evaluation is performed according to the multi-path computing power allocation result to obtain a first performance evaluation result.

[0038] Based on the knowledge distillation method and in combination with the first performance evaluation result, a regeneration configuration of the first group of weak evaluation sub-channels is performed.

[0039] Optionally, construct and train multiple groups of weak evaluation sub-channels. First, define the evaluation objectives, including the quality of resources (clarity, accuracy, completeness, etc.), the effectiveness of resources (learning effect, comprehension, applicability, etc.), and user satisfaction (user feedback, frequency of use, recommendation, etc.).

[0040] Optionally, each group of weak evaluation sub-channels includes a plurality of differentiated weak evaluation sub-channels having a relatively lightweight channel structure, and can efficiently provide evaluation results for the dissemination of a certain construction industry skill. Accordingly, the evaluation accuracy of the decisions of the plurality of weak evaluation sub-channels is relatively low.

[0041] Optionally, in the evaluation subchannel of each resource type, multiple weak evaluation subchannels based on different model structures or parameters are designed to improve the accuracy and robustness of the evaluation. Specifically, the video resource evaluation subchannel includes models based on different structures such as convolutional neural network (CNN), long short-term memory network (LSTM), Transformer, etc. The text resource evaluation subchannel includes models based on different structures such as bag of words model (BoW), TF-IDF, BERT, GPT, etc. The speech resource evaluation subchannel includes models based on different structures such as hidden Markov model (HMM), deep neural network (DNN), convolutional neural network (CNN), etc. The model resource evaluation subchannel includes models based on different structures such as decision tree, random forest, support vector machine (SVM), neural network, etc.

[0042] Through the above method, multiple groups of weak evaluation sub-channels are constructed and trained, which effectively utilizes the differences and diversity between models, avoids the limitations brought by a single model structure and parameter configuration, and is suitable for resource data evaluation scenarios with complex and diverse skills in the construction industry.

[0043] Optionally, the differentiated multiple weak evaluation sub-channels in each group of weak evaluation sub-channels are used to process the same resource data in parallel and give differentiated multiple weak evaluation results. By performing integrated evaluation on the above multiple weak evaluation results, a strong evaluation result with a higher confidence level can be obtained, thereby improving the accuracy and robustness of the evaluation.

[0044] Optionally, according to the multi-path computing power allocation result, a group performance consumption evaluation is performed, including traversing multiple weak evaluation sub-channels in the first group of weak evaluation sub-channels to perform computing power performance evaluation. Specifically, a set of test data sets are prepared, and each weak evaluation sub-channel is run to evaluate the test data set, and the computing resource consumption of each weak evaluation sub-channel (such as GPU usage time, memory consumption, etc.) is recorded; then, the computing resource consumption of all weak evaluation sub-channels is summarized, and the overall computing power consumption of the first group of weak evaluation sub-channels is calculated; then, the overall computing power consumption of the group is compared with the allocated computing power size to determine whether it exceeds the allocated computing power size, and a first performance evaluation result is generated. Among them, the first group of weak evaluation sub-channels is any group among multiple groups of weak evaluation sub-channels.

[0045] Optionally, if the first performance evaluation result shows that the overall computing power consumption of the group is greater than the allocated computing power, model compression processing is performed on multiple weak evaluation sub-channels in the first group of weak evaluation sub-channels to achieve a reconfiguration of the first group of weak evaluation sub-channels.

[0046] Specifically, based on the knowledge distillation method, the regeneration configuration is performed. First, the weak evaluator with the worst performance-to-computing power consumption ratio is selected as the teacher model. Then, based on the output of the teacher model, a smaller student model is trained. The student model can adopt a different structure from the teacher model, but it will learn the predictive distribution of the teacher model. When training the student model, the distillation loss function is used, which combines the traditional cross entropy loss and the soft labels of the teacher model, and introduces a temperature parameter to control the smoothness of the soft labels of the teacher model. Finally, the trained student model is used as a new weak evaluator to replace the original weak evaluator to ensure optimal resource utilization.

[0047] Through the above steps, the performance of the weak evaluator can be effectively improved while optimizing the utilization of computing resources. If the overall computing power consumption exceeds the allocated computing power, the knowledge distillation method can reduce the model complexity and improve the evaluation efficiency while maintaining high performance.

[0048] In some implementations, based on a knowledge distillation method and in combination with the first performance evaluation result, regeneration configuration of the first group of weak evaluation sub-channels is performed, and then the method further includes: Traverse the multiple groups of weak evaluation sub-channels, reconfigure and obtain the multiple groups of weak evaluation sub-channels.

[0049] Based on the resource type distribution characteristics and in combination with a plurality of groups of weak evaluation sub-channels, a multi-path mapper is established.

[0050] A plurality of groups of the weak evaluation sub-channels are connected in parallel and connected to the multi-path mapper to obtain a multi-path data propagation evaluation channel.

[0051] Specifically, all weak evaluation sub-channels of different resource types are traversed. Based on the same method and principle as the above regeneration configuration, for each weak evaluation sub-channel, a teacher model is selected and a corresponding student model is trained, and model compression and regeneration configuration are performed to form multiple groups of regenerated weak evaluation sub-channels.

[0052] Furthermore, according to the distribution characteristics of resource types, appropriate weak evaluation sub-channels are selected for combination. For example, for video resources, weak evaluation sub-channels based on CNN and Transformer models are selected; for text resources, weak evaluation sub-channels based on LSTM and Transformer models are selected. Then, according to the combination of different resource types and weak evaluation sub-channels, a multi-path mapper is established. The multi-path mapper is responsible for mapping different resource types to corresponding weak evaluation sub-channels.

[0053] Optionally, multiple groups of newly configured weak evaluation sub-channels are connected in parallel, and a multipath mapper is connected to the parallel weak evaluation sub-channels, so that different resource types can be allocated to corresponding weak evaluation sub-channels for processing according to resource type distribution characteristics.

[0054] By connecting weak evaluation sub-channels and multi-path mappers in parallel, a multi-path data propagation evaluation channel is formed, which enables dynamic allocation and efficient processing of different types of resource data according to the distribution characteristics of resource types.

[0055] Based on the multi-path data propagation evaluation channel, a propagation evaluation of the multi-path skill resource data set is performed to obtain a propagation evaluation index set.

[0056] Optionally, the multipath skill resource dataset is input into the multipath data propagation evaluation channel. The multipath mapper identifies the type of each resource (such as video, text, image, etc.) and assigns it to the corresponding weak evaluation sub-channel. Each weak evaluation sub-channel performs propagation evaluation on the assigned resource and records the evaluation results.

[0057] Optionally, the propagation evaluation index set is used to reflect the propagation performance of the multi-path skill resource data set. In other words, the propagation evaluation index set reflects the resource learning effectiveness of the multi-path skill resource data set. The higher the propagation evaluation index level in the propagation evaluation index set, the better the teaching and training effect of the multi-path skill resource data set. The propagation evaluation index set includes multiple groups of propagation evaluation indexes, each group of propagation evaluation indexes corresponds to a resource type, including multiple propagation evaluation indexes of multiple weak evaluation sub-channels for the resource type.

[0058] A central value of the propagation evaluation index set is extracted to generate a skill propagation baseline.

[0059] Optionally, the central value may include statistical indicators such as mean, median and mode, which help to form an overall evaluation performance benchmark. The skill propagation baseline is an overall evaluation performance benchmark formed based on the central value of the propagation evaluation index set. It is used to compare the actual evaluation results for skill assessment.

[0060] The skill assessment constraints of the target scenario are obtained, and based on the skill propagation baseline, a graded assessment of the target skill group is performed, and the assessment results are output.

[0061] In some embodiments, obtaining the skill assessment constraints of the target scenario, combining the skill propagation baseline, performing a graded assessment of the target skill group, and outputting the assessment results include: Obtain skill assessment constraints for a target scenario, where the skill assessment constraints include multiple groups of assessment constraints.

[0062] Based on the skill assessment constraints and according to the skill propagation baseline, a multi-path assessment channel is constructed, wherein the multi-path assessment channel includes a simulation assessment channel and a practical assessment channel.

[0063] Apply the multi-path assessment channel to assess the target skill group and output the assessment result.

[0064] Optionally, skill assessment constraints of the target scenario are obtained, and skill assessment constraints are standards and conditions for evaluating the skill level in the target skill group. For example, for welding skills, the corresponding skill assessment constraints include weld firmness, weld surface quality, number of welding defects, safe operation, welding efficiency, familiarity with relevant knowledge, etc.

[0065] Optionally, a multi-path assessment channel is constructed based on the skill assessment constraints and skill propagation baseline, including a simulation assessment channel and a practical assessment channel. The simulation assessment channel is constructed based on VR / MR simulation technology and is used to simulate actual scenarios to assess construction industry skills. The practical assessment channel is constructed based on a preset practical assessment plan and is used for actual operations and drills, and the skills are assessed based on the results of the operations and drills.

[0066] Exemplarily, the target skill group is assessed in a graded manner based on the assessment results. If the performance is excellent in all assessment constraints and exceeds the mean and median of the skill propagation baseline, the assessment result is excellent. If the performance is good in most assessment constraints and is close to or exceeds the median of the skill propagation baseline, the assessment result is good. If the performance is qualified in the basic assessment constraints and meets the minimum requirements of the skill propagation baseline, the assessment result is qualified. If the basic assessment constraints and the minimum requirements of the skill propagation baseline are not met, the assessment result is unqualified.

[0067] In some implementations, based on the multiple groups of assessment constraints and according to the skill propagation baseline, constructing a multi-path assessment channel includes: Establish an association relationship between the multiple groups of assessment constraints and the skill propagation baseline.

[0068] Based on the skill propagation baseline, the simulation assessment channel is constructed, and based on the multiple groups of assessment constraints, the practical assessment channel is constructed.

[0069] The simulation assessment channel and the practical assessment channel are connected to obtain a multi-path assessment channel.

[0070] Optionally, establish the relationship between multiple groups of assessment constraints and skill diffusion baselines. First, clarify the specific content and evaluation criteria of each skill assessment constraint; then, match each assessment constraint with the corresponding indicator in the skill diffusion baseline; finally, record the relationship between each assessment constraint and the baseline indicator for comparison and analysis during the evaluation process.

[0071] The simulation assessment channel is a channel for assessing candidates’ skills by simulating actual scenarios and automated tools. Specifically, to build a simulation assessment channel, first select appropriate simulation tools, such as VR and MR, and then configure the simulation environment and test cases based on the skill propagation baseline.

[0072] Among them, the practical assessment channel is a channel for evaluating candidates' skills through actual operations and drills. Specifically, building a practical assessment channel includes: designing specific practical tasks and projects based on assessment constraints; and clarifying the assessment standards and scoring rules for each practical task.

[0073] Specifically, the simulation assessment channel built based on the skill propagation baseline is used to assess the skill level of construction industry practitioners under ideal conditions, which is a preliminary assessment of construction industry skills. The practical assessment channel is an advanced assessment based on assessment constraints and is used to assess the skill level of construction industry practitioners in a practical environment.

[0074] Optionally, a multi-path assessment channel is obtained by connecting the simulation assessment channel and the practical assessment channel through a discriminator. The discriminator is used to determine whether the preliminary assessment result of the simulation assessment channel meets the skill propagation baseline. If so, an advanced assessment is performed based on the practical assessment channel.

[0075] Optionally, if the preliminary assessment results of the simulation assessment channel do not meet the skill transfer baseline, the construction industry practitioners are guided to the simulation assessment channels corresponding to other skills in the target skill group for extended assessment until the entire target skill group or the practical assessment channel of a certain construction industry skill is traversed and the assessment result is passed.

[0076] In summary, the multi-path skill assessment method for the construction industry provided by the present invention has the following technical effects: Through the interactive skill resource database, the multi-path skill resource data set corresponding to the target skill group is extracted, wherein the multi-path skill resource data set includes the multi-dimensional skill resource data of multiple skills in the target skill group; a multi-path data propagation evaluation channel is constructed, and the multi-path data propagation evaluation channel includes multiple groups of weak evaluation sub-channels; based on the multi-path data propagation evaluation channel, the multi-path skill resource data set is propagated and evaluated to obtain the propagation evaluation index set; the centralized value of the propagation evaluation index set is extracted to generate the skill propagation baseline; the skill assessment constraints of the target scenario are obtained, and the target skill group is graded and evaluated in combination with the skill propagation baseline, and the assessment results are output. Thus, the technical effect of accurately evaluating the comprehensive quality and helping to give full play to the individual advantages is achieved.

[0077] Embodiment 2

[0078] Based on the same concept as the multi-path skill assessment method for the construction industry in the above embodiment, Figure 2 As shown, the present application also provides a multi-path skill assessment system for the construction industry, the system comprising: A resource acquisition module 11 is used to interact with a skill resource database and extract a multi-path skill resource data set corresponding to a target skill group, wherein the multi-path skill resource data set includes multi-dimensional skill resource data of multiple skills in the target skill group; A propagation assessment configuration module 12, used to construct a multipath data propagation assessment channel, wherein the multipath data propagation assessment channel includes a plurality of groups of weak assessment sub-channels; A propagation evaluation module 13, configured to perform a propagation evaluation of the multi-path skill resource data set based on the multi-path data propagation evaluation channel to obtain a propagation evaluation index set; A propagation base acquisition module 14 is used to extract the central value of the propagation evaluation index set to generate a skill propagation baseline; The multi-path assessment module 15 is used to obtain the skill assessment constraints of the target scenario, conduct a graded assessment of the target skill group in combination with the skill propagation baseline, and output the assessment results.

[0079] The resource acquisition module 11 includes: A target acquisition unit, used to interact with target scenarios and acquire target skill sets; An association division unit, used for association analysis of the target skill set to generate a plurality of associated skill groups; The resource extraction unit is used to perform random extraction among the multiple associated skill groups to obtain the target skill group, and use the target skill group as an index constraint to traverse the skill resource database and extract the multi-path skill resource data set.

[0080] Furthermore, the propagation evaluation configuration module 12 includes: A category feature analysis unit, used to analyze the multi-path skill resource data set to obtain resource category distribution features; A computing power allocation unit, used to evaluate the computing power of the target scenario and perform multi-path computing power allocation based on the resource type distribution characteristics; A multipath configuration unit is used to configure the multipath data propagation evaluation channel based on the multipath computing power allocation result.

[0081] In some implementations, the propagation evaluation configuration module 12 includes: An initial configuration unit, configured to construct and train multiple groups of weak evaluation sub-channels based on resource type distribution characteristics, wherein the multiple groups of weak evaluation sub-channels correspond to multiple resource types; An initial configuration evaluation unit, configured to extract a first group of weak evaluation sub-channels, perform group performance consumption evaluation according to the multi-path computing power allocation result, and obtain a first performance evaluation result; A configuration regeneration unit is used to regenerate the configuration of the first group of weak evaluation sub-channels based on the knowledge distillation method and in combination with the first performance evaluation result.

[0082] In some implementations, the propagation evaluation configuration module 12 further includes: A traversal regeneration unit, used for traversing the plurality of groups of the weak evaluation sub-channels, regenerating and configuring to obtain the plurality of groups of the weak evaluation sub-channels; A path mapping unit, configured to establish a multi-path mapper based on the resource type distribution characteristics and in combination with a plurality of groups of weak evaluation sub-channels; The multipath integration unit is used to connect multiple groups of the weak evaluation sub-channels in parallel, connect the multipath mapper, and obtain a multipath data propagation evaluation channel.

[0083] In some embodiments, the multi-path assessment module 15 includes: A constraint acquisition unit, used to acquire skill assessment constraints of a target scenario, wherein the skill assessment constraints include multiple groups of assessment constraints; A multi-channel construction unit, configured to construct a multi-path assessment channel based on the skill assessment constraint and the skill propagation baseline, wherein the multi-path assessment channel includes a simulation assessment channel and a practical assessment channel; The skill assessment unit is used to apply the multi-path assessment channel to assess the target skill group and output the assessment result.

[0084] In some implementations, the multi-channel construction unit in the multi-path assessment module 15 includes: A baseline association unit, used to establish an association relationship between the multiple groups of assessment constraints and the skill propagation baseline; An assessment channel construction unit, configured to construct the simulation assessment channel based on the skill propagation baseline, and to construct the practical assessment channel based on the multiple groups of assessment constraints; A sequential connection unit is used to connect the simulation assessment channel and the practical assessment channel to obtain a multi-path assessment channel.

[0085] It should be understood that the embodiments mentioned in this specification focus on their differences from other embodiments. The specific embodiments in the aforementioned embodiment one are also applicable to the multi-path skill assessment system for the construction industry described in embodiment two. For the sake of brevity of the specification, they will not be further elaborated here.

[0086] It should be understood that the embodiments disclosed in this application and the above description can enable those skilled in the art to use this application to implement this application. At the same time, this application is not limited to the above-mentioned embodiments, and obvious modifications, combinations and substitutions of the embodiments mentioned in this application also fall within the scope of protection of this application.

Claims

1. A multi-path skills assessment method for the construction industry, characterized in that: The method comprises: Interactive skill resource database, extracting a multi-path skill resource data set corresponding to the target skill group, wherein the multi-path skill resource data set includes multi-dimensional skill resource data of multiple skills in the target skill group; Constructing a multipath data propagation evaluation channel, wherein the multipath data propagation evaluation channel includes a plurality of groups of weak evaluation sub-channels; Performing a propagation evaluation of the multi-path skill resource data set based on the multi-path data propagation evaluation channel to obtain a propagation evaluation index set; Extracting a central value of the propagation evaluation index set to generate a skill propagation baseline; The skill assessment constraints of the target scenario are obtained, and based on the skill propagation baseline, a graded assessment of the target skill group is performed, and the assessment results are output.

2. The method according to claim 1, characterized in that Interactive skill resource database, extracting the multi-path skill resource data set corresponding to the target skill group, including: Interact with target scenarios to acquire target skill sets; Associatively analyzing the target skill set to generate multiple associated skill groups; Random extraction is performed among the multiple associated skill groups to obtain the target skill group, and the skill resource database is traversed with the target skill group as an index constraint to extract the multi-path skill resource data set.

3. The method according to claim 2, characterized in that Construct a multi-path data propagation assessment channel, including: Analyze the multi-path skill resource data set to obtain resource type distribution characteristics; Conduct computing power evaluation for the target scenario, and perform multi-path computing power allocation based on the resource type distribution characteristics; Based on the multi-path computing power allocation result, the multi-path data propagation evaluation channel is configured.

4. The method according to claim 3, characterized in that Based on the multi-path computing power allocation result, configuring the multi-path data propagation evaluation channel includes: Based on the resource type distribution characteristics, multiple groups of weak evaluation sub-channels are constructed and trained, and the multiple groups of weak evaluation sub-channels correspond to multiple resource types; Extracting a first group of weak evaluation sub-channels, performing group performance consumption evaluation according to the multi-path computing power allocation result, and obtaining a first performance evaluation result; Based on the knowledge distillation method and in combination with the first performance evaluation result, a regeneration configuration of the first group of weak evaluation sub-channels is performed.

5. The method according to claim 4, characterized in that Based on the knowledge distillation method and in combination with the first performance evaluation result, the first group of weak evaluation sub-channels are regenerated and configured, and then, the method further includes: Traversing the plurality of groups of the weak evaluation sub-channels, reconfiguring and acquiring the plurality of groups of the weak evaluation sub-channels; Based on the resource type distribution characteristics, a multipath mapper is established in combination with a plurality of groups of weak evaluation sub-channels; A plurality of groups of the weak evaluation sub-channels are connected in parallel and connected to the multi-path mapper to obtain a multi-path data propagation evaluation channel.

6. The method according to claim 5, characterized in that Obtain the skill assessment constraints of the target scenario, combine the skill propagation baseline, perform graded assessment of the target skill group, and output the assessment results, including: Acquire skill assessment constraints of a target scenario, wherein the skill assessment constraints include multiple groups of assessment constraints; Based on the skill assessment constraints and the skill propagation baseline, a multi-path assessment channel is constructed, wherein the multi-path assessment channel includes a simulation assessment channel and a practical assessment channel; Apply the multi-path assessment channel to assess the target skill group and output the assessment result.

7. The method according to claim 6, characterized in that Based on the multiple groups of assessment constraints and according to the skill propagation baseline, a multi-path assessment channel is constructed, including: Establishing a correlation relationship between the multiple groups of assessment constraints and the skill propagation baseline; Based on the skill propagation baseline, construct the simulation assessment channel, and based on the multiple groups of assessment constraints, construct the practical assessment channel; The simulation assessment channel and the practical assessment channel are connected to obtain a multi-path assessment channel.

8. A multi-path skill assessment system for the construction industry, characterized in that: The system is used to execute the multi-path skill assessment method for the construction industry according to any one of claims 1 to 7, and the system comprises: A resource acquisition module, the resource acquisition module is used to interact with the skill resource database to extract a multi-path skill resource data set corresponding to the target skill group, wherein the multi-path skill resource data set includes multi-dimensional skill resource data of multiple skills in the target skill group; A propagation evaluation configuration module, wherein the propagation evaluation configuration module is used to construct a multi-path data propagation evaluation channel, wherein the multi-path data propagation evaluation channel includes a plurality of groups of weak evaluation sub-channels; A propagation evaluation module, the propagation evaluation module is used to perform a propagation evaluation of the multi-path skill resource data set based on the multi-path data propagation evaluation channel to obtain a propagation evaluation index set; A propagation base acquisition module, the propagation base acquisition module is used to extract the central value of the propagation evaluation index set to generate a skill propagation baseline; A multi-path assessment module is used to obtain the skill assessment constraints of the target scenario, combine the skill propagation baseline, conduct a graded assessment of the target skill group, and output the assessment results.

Images