Teaching test method and system for vocational education

Abstract

The invention relates to a teaching test method and system for vocational education, and belongs to the technical field of teaching test systems. Comprising the steps of building a feedback participation framework of student education and teaching, packaging training project homework and related information of test skill points into a transaction, and sending the transaction to an alliance block chain network; encrypting the answer data and the generated answer information, creating a transaction, and sending the encrypted data to the cloud platform; scoring and screening the answer data based on the reference answers through the received student answer data; packaging the screened answer data into a to-be-verified block for consensus; and evaluating the mastering degree of the skill points of the students through the check score and the decision result of the consensus process. According to the technical scheme of the invention, the interactive feedback platform is constructed to instantly release the training project operation of the skill points, so that the efficient and timely testing is realized, and the knowledge of the mastering degree of the skill points is further improved in combination with the answering data score and mutual evaluation inspection mode.

Description

Technical Field

[0001] The invention belongs to the technical field of teaching test systems, and in particular relates to a teaching test method and system for vocational education. Background Art

[0002] With the rapid development of information technology, intelligent education has gradually become an important direction of vocational education. Vocational education aims to provide students with practical skills training so that they can better adapt to social needs.

[0003] However, the traditional vocational education teaching model has many deficiencies in the feedback mechanism, which limits the improvement of teaching effect. At present, the method of testing students' mastery of knowledge points is still mainly based on tests and examinations. The indicator for measuring the mastery of knowledge points is still the score of the corresponding knowledge points. The description of the mastery of knowledge points is also limited to the analysis of the mastery of knowledge points. The direct basis is still the recognition of the right and wrong questions, that is, the score rate is calculated. Although the score is a direct display of the answering situation of the test paper, the learning level of two students with the same score is not necessarily exactly the same. The score cannot directly observe the learning level of the students, and it is also impossible to directly judge the students' mastery of knowledge points. At the same time, tests and examinations are conducted in stages, and feedback cannot be flexibly realized, which affects the teacher's teaching plan. Summary of the invention

[0004] In order to solve the above-mentioned problems existing in the prior art, the present invention provides a teaching test method and system for vocational education, which realizes efficient and timely testing by constructing an interactive feedback platform to instantly publish practical training project assignments of skill points, and further enhances the understanding of the mastery of skill points by setting questions for skill points in a fine-grained manner, combining answer data scoring and mutual evaluation and verification.

[0005] The purpose of the present invention can be achieved by the following technical solutions: A first aspect of the present disclosure provides a teaching test method for vocational education, comprising the following steps: S1. Create an interactive feedback platform: build a feedback participation framework for student education and teaching based on the alliance chain, which includes the user end, alliance blockchain and cloud platform; S2. Release the practical training project assignment: package all the data of the practical training project assignment and related information of the test skill points into a transaction through smart contracts, and send the transaction to the alliance blockchain network; S3. Participant answer submission: Students submit answer data in different data formats on the interactive feedback platform, encrypt the answer data and generated answer information, create a transaction, and send the encrypted data to the cloud platform; S4, cloud platform review and scoring: score the received student answer data based on the reference answers, and after obtaining the scores, send them back to the alliance chain block network to screen the answer data; S5. Consensus of Participants’ Answers: Set a score threshold to filter the obtained scores, and package the filtered answer data into blocks to be verified. All students participating in the practical training project work will serve as a consensus group to reach consensus on the blocks; S6. Assessment of students’ skills: The students’ mastery of the skills tested in the practical training project will be assessed by reviewing and grading the students’ answers and the decision-making results of the consensus process; The student skills assessment also includes: The review scores of students' answer data are used as the basic scores. The number of correct decisions made by students for each skill point during the consensus process of the answer data is counted as the weighted score to calculate the students' comprehensive scores. Among them, the correctness of the decision on the skill point is based on the decision results for the skill point during the teacher consensus process.

[0006] Furthermore, the user terminal is used to interact with the alliance blockchain network through an application program interface and access data on the alliance blockchain through smart contracts; The alliance blockchain is used to authenticate the user information of the user end and save the process record of students participating in the practical training project assignments through automatic execution of smart contracts, and send the answer data to the cloud platform; The cloud platform is used to build a multimodal data conversion model, convert the received student answer data into text data and score it, and then send the scoring results to the alliance blockchain.

[0007] Furthermore, the smart contract includes the following rules: User information authentication: Design an authentication mechanism to verify the identity of the user. After the authentication is passed, the student information will be stored in the alliance blockchain; Practical training project assignment release: teachers or schools serve as master nodes to release practical training project assignments. The released information includes project ID, assignment content and deadline, and is recorded on the alliance blockchain; Receiving and automatically scoring answer data: After the user submits the answer, the answer data will be sent to the alliance blockchain and marked as "pending scoring"; at the same time, the answer data will be sent to the cloud platform for automatic scoring, and the scoring results will be returned and recorded on the alliance blockchain to update the status.

[0008] Furthermore, the smart contract also includes rules: Consensus mechanism: The answer data are screened out according to the scoring results and score thresholds for consensus. The participants of the practical training project work serve as a consensus group to review each answer data and record the decision results of each skill point; New block generation: During the consensus process, a new block is generated for the answer data with the highest support. The new block contains the signatures of the supporters and the consensus result, and then the new block is broadcast to all nodes of the blockchain for data synchronization.

[0009] Furthermore, the cloud platform review and scoring includes the steps of: Multimodal data conversion: distinguish the answer data, convert the voice data and video data to generate text descriptions, where the voice data outputs voice text through the voice recognition model, and the video data first captures the timing information of continuous frames in the video through the timing model, and then obtains the text description of the student's behavior sequence through the behavior recognition model.

[0010] Furthermore, the cloud platform review and scoring further includes the following steps: Answer data feature extraction: The converted text of the student's answer data and the reference answer are input into the BERT model for feature extraction. The text is processed through a bidirectional encoder to generate context information, and the last layer output of the BERT model is extracted as the feature vector of the answer data. Scoring model construction: Build a multi-layer perceptron network to process the feature vectors generated by BERT, introduce a gated recurrent unit model to process sequence data, capture the context information of the answer data, and fuse the output results of MLP and GRU to form the final scoring model; Obtaining answer scores: The MLP part calculates the textual implication relationship between the answer data and the reference answer to evaluate the relevance of the answer; the GRU part calculates the similarity between the answer data and the standard answer to evaluate the logic and coherence of the answer; the textual implication score and the textual similarity score are weightedly integrated to obtain the final score of the answer data.

[0011] Furthermore, the participants' answer consensus includes the following steps: Select a consensus algorithm to initiate verification of the new block. Each node judges the correctness of the description of the skill point based on the answer data and gives a decision result. The decision result based on each skill point serves as a support vote for each answer data. The node collects the voting results of all participants, selects the block with the highest support according to the voting results to reach a consensus, and calculates the hash value of the block; the support includes the total number of times the consensus group supports all skill points in the answer data; The consensus block is added to the alliance blockchain and broadcast to other nodes in the network. Other nodes verify the validity of the new block and update their own ledgers.

[0012] Furthermore, the calculation formula for calculating the comprehensive score of the student is as follows: F i=S i +(W×C i ); In the formula, F i For Students i The comprehensive score, S i For Students i The basic score, W is the weight coefficient, C i For Students i The number of correct decisions.

[0013] A second aspect of the present disclosure provides a teaching test system for vocational education, which is used to implement a teaching test method for vocational education as described above, including a user management module, a homework publishing and answering module, an answer data scoring module and a skill mastery assessment module; The user management module is used to support the registration and identity authentication of students and teachers, ensure the legitimacy of users, assign permissions to different roles, manage personal information, and maintain user access to system functions; The homework publishing and answering module is used for teachers or schools to publish practical training project homework online, and students answer and submit them through user-end devices. At the same time, answer information is generated and encrypted and stored in the alliance blockchain network.

[0014] As a preferred technical solution of the present invention, the answer data scoring module is used to automatically score and quickly feedback the answer data submitted by the students through the multimodal answer data scoring model constructed by the cloud platform, and continuously update the scoring model; The updating of the scoring model includes the following steps: According to the teaching courses, test skill points and reference answers, add labels to the recorded practical videos or answer data, and train the multimodal conversion model through the continuously improved training data set; Collect the answer data obtained by consensus to update the training data set to train the scoring model and optimize the model structure and parameters; The skill mastery assessment module is used to assess students' mastery of the skill points in the practical training project assignment test by collecting students' answer data, reviewing the scoring results and the decision results of the consensus process on the skill points, and calculating the comprehensive score of each student.

[0015] The beneficial effects of the present invention are: The present invention first creates an interactive feedback platform for obtaining student teaching feedback information based on blockchain, and uses the tamper-proof and distributed characteristics of blockchain technology to fully record the student's answer data and the teacher's practical training project homework release process, and then automatically scores the answer data through the cloud platform review and scoring, and selects part of the answer data based on the obtained score for consensus, and selects the answer data with the highest support to generate a new block for sharing, and finally evaluates the student's mastery of skill points through the student's answer data and consensus records, realizes feedback on teaching effects through timely testing of educational and teaching skill points, and strengthens learning and understanding of the skill points released for testing through consensus.

[0016] The present invention establishes a testing platform based on blockchain. Through the process that teachers upload skill point test assignments to the chain in time and score and reach consensus on the answer data, the effect of timely understanding the students' learning status is achieved. At the same time, based on the multimodal scoring system, it provides support for submitting assignments in multiple ways, meets the needs of vocational skill testing, and improves the efficiency of interaction between students and teachers. Before the consensus process, all answer data are first scored based on reference answers to establish a comprehensive scoring basis, and then consensus is reached based on the student's answer score screening part, which improves the efficiency of consensus. At the same time, in the consensus process, the decision results are obtained through fine-grained division of data, and the students' understanding of skill points is understood in detail, which improves the feedback effect and the effectiveness of the comprehensive scoring. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to facilitate understanding by those skilled in the art, the present invention is further described below with reference to the accompanying drawings.

[0018] Figure 1 A flow chart of a teaching test method for vocational education provided by an embodiment of the present invention; Figure 2 A schematic diagram of the structure of a feedback participation framework provided by an embodiment of the present invention; Figure 3 A schematic diagram of a cloud platform review and scoring process provided by an embodiment of the present invention; Figure 4 A structural diagram of a teaching and testing system for vocational education provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0019] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the specific implementation methods, structures, features and effects of the present invention are described in detail below in conjunction with the accompanying drawings and preferred embodiments.

[0020] Based on the problems of unsatisfactory score feedback effect and limited flexibility, the present invention constructs an interactive feedback platform in this embodiment, which realizes efficient and timely testing by instantly publishing practical training project assignments of skill points, and tests skill points by setting questions in a fine-grained manner, combining answer data scoring and mutual evaluation and verification, thereby further improving the understanding of the mastery of skill points.

[0021] Specifically, the teaching test method and system for vocational education of the present invention are described in detail through the following content: This embodiment provides a teaching test method for vocational education, such as Figure 1 As shown, the following steps are included: S1. Create an interactive feedback platform: Build a feedback participation framework for student education and teaching based on the alliance chain. The feedback participation framework is as follows: Figure 2 As shown, it includes the user end, alliance blockchain and cloud platform; The user terminal is used to interact with the alliance blockchain network through an application programming interface (API) and access the data on the alliance blockchain through smart contracts.

[0022] The alliance blockchain is used to authenticate the user information of the user end and save the process record of students participating in the practical training project assignments through automatic execution of smart contracts, and send the answer data to the cloud platform; The cloud platform is used to build a multimodal data conversion model, convert the received student answer data into text data and score it, and then send the scoring results to the alliance blockchain.

[0023] It should be noted that by building an interactive feedback platform based on the alliance blockchain, a foundation is provided for students' skill mastery testing. The blockchain technology has the characteristics of decentralization, distributed storage and non-tamperability, which solves the problem of limited development of shared learning platforms under the traditional centralized management model due to lack of trust. Students can test and evaluate the flexibility of knowledge points to timely understand the effect of education and teaching. Students use commonly used small mobile devices on the learning end as nodes of the alliance blockchain to realize information sharing and interaction.

[0024] Among them, the smart contract includes the following rules: User information authentication: Design an authentication mechanism to verify the identity of the user. After the authentication is passed, the student information (such as student ID, name, authentication time, etc.) will be stored in the alliance blockchain.

[0025] Publishing of practical training project assignments: Teachers or schools act as master nodes to publish practical training project assignments. The published information includes project ID, assignment content, deadline, etc., and is recorded on the alliance blockchain.

[0026] Receiving answer data and automatically scoring: After the user submits the answer, the answer data is sent to the alliance blockchain and marked as "pending scoring"; at the same time, the answer data is sent to the cloud platform for automatic scoring, and the scoring results are returned and recorded on the alliance blockchain to update the status; Consensus mechanism: The answer data are screened out according to the scoring results and score thresholds for consensus. The participants of the practical training project work serve as a consensus group to review each answer data and record the decision results of each skill point; New block generation: During the consensus process, a new block is generated for the answer data with the highest support. The new block contains the signatures of the supporters and the consensus result, and then the new block is broadcast to all nodes of the blockchain for data synchronization.

[0027] S2. Release practical training project assignments: Package all data of practical training project assignments and related information (such as assignment ID, deadline, etc.) for testing skill points into a transaction through smart contracts, and send the transaction to the alliance blockchain network.

[0028] It should be noted that in order to ensure the effectiveness of feedback, teachers need to publish test assignments immediately and set questions based on skill points to achieve fine-grained testing. Through a small number of high-frequency tests, timely feedback on education and teaching can be achieved, which facilitates teachers to rationally plan and adjust teaching plans.

[0029] S3. Participants submit answers: Students submit answer data in different data formats on the interactive feedback platform, encrypt the answer data and the answer information generated by the platform (such as student ID, submission time, assignment ID, etc.), create transactions, and send the encrypted data to the cloud platform.

[0030] S4, cloud platform review and scoring: obtain the student's answer data by decrypting the received data, score the answer data based on the reference answer, and then send the score back to the alliance chain block network to filter the answer data. Figure 3 As shown, the scoring includes the following steps: S41. Multimodal data conversion: distinguish the answer data, convert the voice data and video data to generate text descriptions, where the voice data outputs voice text through the voice recognition model, and the video data captures the timing information of continuous frames in the video through the timing model (such as LSTM or 3D CNN), and then obtains the text description of the student's behavior sequence through the behavior recognition model; It is understandable that the main purpose of the above steps is to convert multimodal data into text for scoring. The collection of multimodal data is to achieve direct feedback on the teaching effect of vocational education, and to reduce the examination limitations of practical training project assignments by satisfying the scoring of multimodal answer data.

[0031] Regarding student behavior recognition, during model training, practical behaviors are labeled for different educational courses to improve the recognition accuracy of the model's professional actions, and recognition and description are performed according to time-series video frames to record the behavioral norms of the operation process, thereby providing accurate data for scoring.

[0032] S42. Feature extraction of answer data: Input the converted text and reference answers of the student answer data into the BERT (Bidirectional Encoder Representations from Transformers) model for feature extraction. The text is processed by the bidirectional encoder to generate context information, and the last layer output of the BERT model is extracted as the feature vector of the answer data. S43, Scoring model construction: Construct a multi-layer perceptron (MLP) network to process the feature vectors generated by BERT to calculate the textual implications between the answer data and the reference answer; The Gated Recurrent Unit (GRU) model is introduced to process sequence data and capture the contextual information of the answer data to evaluate the fluency and logic of the answer.

[0033] The output results of MLP and GRU are fused to form the final scoring model.

[0034] It should be noted that the multilayer perceptron is a basic artificial neural network (ANN) model that can realize nonlinear mapping from input state space to output state space. In the practical training project assignment, this embodiment collects teaching effect feedback by students describing the practical process steps or their feelings about the practical training operation. Therefore, the answer data is sequence data. The sequence feature is extracted through the gated recurrent unit model. The GRU model can not only solve the problem of gradient disappearance and capture long-term dependencies, but also has fewer parameters and is more efficient in training and reasoning.

[0035] S44. Obtain answer scores: Calculate the textual implication relationship between the answer data and the reference answer through the MLP part to evaluate the relevance of the answer; Calculate the similarity between the answer data and the standard answer through the GRU part to evaluate the logic and coherence of the answer; Perform weighted fusion of the textual implication score and the textual similarity score to obtain the final score of the answer data.

[0036] It is understandable that the above scoring process, based on the BERT model, integrates the perceptron and gated recurrent unit to establish an MLP-GRU text question automatic scoring model. This model simultaneously considers the implication relationship and similarity between the reference answer and the student's answer, and realizes a multi-angle comprehensive evaluation of the student's answer. First, the BERT model is used to extract the features of the student's answer, providing rich semantic information for the subsequent scoring process. Then the perceptron and GRU models are introduced, respectively combined with the BERT output to calculate the text implication and text similarity scores between the standard answer, to ensure that the fluency and logic of the answer are considered. Finally, the text implication score and text similarity score are integrated to obtain the final scoring result.

[0037] It should be noted that in order to meet the requirements of vocational education and teaching effectiveness testing, a behavior recognition model is set up on the cloud platform to identify the practical operation process recorded in the student video, thereby completing the collection of student feedback data. Then, the student answer data is scored based on the text description of the practical operation process based on the reference answer, so as to realize the degree of student skill mastery. At the same time, a cloud platform scoring model is set up to adapt to the instant test method, which reduces the workload of teachers and improves the efficiency of the system in realizing teaching effect feedback.

[0038] S5. Participants’ consensus on answers: Set a score threshold to filter the obtained scores, and package the filtered answer data into blocks to be verified. All students and teachers participating in the practical project work will serve as a consensus group to reach a consensus on the blocks, including: Select a consensus algorithm to initiate verification of the new block. Each node determines the correctness of the description of the skill point based on the answer data and gives a decision result (such as support or not). The decision result based on each skill point serves as a support vote for each answer data. The node collects the voting results of all participants, selects the block with the highest support based on the voting results to reach a consensus, and calculates the hash value of the block; the support includes the sum of the number of times the consensus group supports all skill points in the answer data.

[0039] The blocks reached after consensus are added to the alliance blockchain and broadcast to other nodes in the network. Other nodes verify the validity of the new blocks and update their own ledgers to keep the entire network status consistent.

[0040] It should be noted that the purpose of the cloud platform review and scoring is to first conduct a standardized and unified evaluation of the answer data based on the reference answers, and secondly to screen the consensus data to improve the efficiency of the consensus. During the consensus process, the answer data is finely decomposed based on key skill points, and decisions are made for each skill point. Finally, the consensus result is selected based on the total support. The new block obtained through the consensus mechanism contains the answer data with the highest support. Students can access it through their own nodes, achieving different enhanced understandings of the knowledge points tested.

[0041] S6. Assessment of student skills mastery: The assessment is conducted by collecting the review scores of the students’ answers and the decision results of the consensus process to assess the students’ mastery of the skills tested in the practical training project assignments, including the following steps: S61. Using the review scores of the students' answer data as the basic scores, and counting the number of correct decisions made by the students for each skill point during the consensus process of the answer data; It should be noted that the judgment of whether the decision on skill points is correct or not is based on the decision results on skill points during the teacher consensus process. If the consensus makers make a correct judgment, it is supported; if the judgment is wrong, it is not supported.

[0042] S62. Calculate the comprehensive score of the student. The calculation formula is as follows: F i =S i +(W×C i ); In the formula, F i For Students i The comprehensive score, S i For Students i The basic score, W is the weight coefficient, C i For Students i The number of correct decisions.

[0043] It should be noted that the weight coefficient is an adjustable parameter used to set the degree of influence of the number of correct decisions on the comprehensive score. In order to ensure that the comprehensive scores of all students are within a uniform range (such as 0 to 100), normalization can be performed. The above-mentioned assessment of students' mastery of skill points includes, on the one hand, the students' own answers to the skill point training projects, which directly tests the students' mastery of skill point knowledge. On the other hand, it includes the students' fine-grained support results for other answer results, which tests the students' cognition of the details of the entire practical process from different angles, further improving the system's deeper understanding of the students' mastery.

[0044] This embodiment also provides a teaching test system for vocational education, such as Figure 4As shown, it includes a user management module, a homework publishing and answering module, an answering data scoring module and a skill mastering assessment module; The user management module is used to support the registration and identity verification of students and teachers, ensure the legitimacy of users, assign permissions to different roles, manage personal information, and maintain user access to system functions.

[0045] The homework publishing and answering module is used for teachers or schools to publish practical training project homework online, and students answer and submit them through user-end devices. At the same time, answer information is generated and encrypted and stored in the alliance blockchain network.

[0046] The answer data scoring module is used to automatically score the answer data submitted by students and quickly feedback the scores through the multimodal answer data scoring model built by the cloud platform, and continuously update the scoring model. The updating of the scoring model includes the following steps: According to the teaching courses, test skill points and reference answer descriptions, labels are added to the recorded practical videos or answers, and the multimodal conversion model is trained through a continuously improved training data set to improve the accuracy of multimodal data conversion.

[0047] The answer data obtained by consensus is collected to update the training data set for scoring model training, further optimize the model structure and parameters, and improve the scoring performance.

[0048] It is understandable that through targeted annotation of multimodal data conversion, it is ensured that during the conversion process, the words of the reference answer are used for text description, so as to ensure the provision of accurate conversion answers in subsequent scoring. At the same time, the scoring model is optimized by utilizing the answer data obtained by consensus. This process is equivalent to manual identification and screening of training data, thereby improving the scoring performance of the model through training.

[0049] The skill mastery assessment module is used to assess students' mastery of practical training project assignment test skill points and their ability to apply professional skills by collecting student answer data, reviewing scoring results and the decision results of the consensus process on skill points, and calculating a comprehensive score for each student.

[0050] The present invention first creates an interactive feedback platform for obtaining student teaching feedback information based on blockchain, and uses the tamper-proof and distributed characteristics of blockchain technology to fully record the student's answer data and the teacher's practical training project homework release process, and then automatically scores the answer data through the cloud platform review and scoring, and selects part of the answer data based on the obtained score for consensus, and selects the answer data with the highest support to generate a new block for sharing, and finally evaluates the student's mastery of skill points through the student's answer data and consensus records, realizes feedback on teaching effects through timely testing of educational and teaching skill points, and strengthens learning and understanding of the skill points released for testing through consensus.

[0051] The present invention establishes a testing platform based on blockchain. Through the process that teachers upload skill point test assignments to the chain in time and score and reach consensus on the answer data, the effect of timely understanding the students' learning status is achieved. At the same time, based on the multimodal scoring system, it provides support for submitting assignments in multiple ways, meets the needs of vocational skill testing, and improves the efficiency of interaction between students and teachers. Before the consensus process, all answer data are first scored based on reference answers to establish a comprehensive scoring basis, and then consensus is reached based on the student's answer score screening part, which improves the efficiency of consensus. At the same time, in the consensus process, the decision results are obtained through fine-grained division of data, and the students' understanding of skill points is understood in detail, which improves the feedback effect and the effectiveness of the comprehensive scoring.

[0052] The above description is only a preferred embodiment of the present invention and does not limit the present invention in any form. Although the present invention has been disclosed as a preferred embodiment as above, it is not used to limit the present invention. Any technical personnel in this field can make some changes or modify the technical contents disclosed above into equivalent embodiments without departing from the scope of the technical solution of the present invention. However, any brief modifications, equivalent changes and modifications made to the above embodiments based on the technical essence of the present invention without departing from the content of the technical solution of the present invention are still within the scope of the technical solution of the present invention.

Claims

1. A teaching test method for vocational education, characterized in that: The following steps are involved: Create an interactive feedback platform: build a feedback participation framework for student education and teaching based on the alliance chain, which includes the user end, alliance blockchain and cloud platform; Release practical training project assignments: Package all data of practical training project assignments and related information for testing skill points into one transaction through smart contracts, and send the transaction to the alliance blockchain network; Participants submit answers: Students submit answer data in different data formats on the interactive feedback platform, encrypt the answer data and generated answer information, create transactions, and send the encrypted data to the cloud platform; Cloud platform review and scoring: Scoring is performed based on the received student answer data and the reference answers. After obtaining the scores, they are sent back to the alliance chain block network to screen the answer data. Participants' answer consensus: Set a score threshold to filter the obtained scores, package the filtered answer data into blocks to be verified, and all students participating in the practical training project work will serve as a consensus group to reach consensus on the blocks; Assessment of student skills mastery: The degree of students' mastery of the skills tested in the practical training project assignments is assessed by reviewing and grading the students' answer data and the decision-making results of the consensus process; The student skills assessment also includes: The review scores of students' answer data are used as the basic scores. The number of correct decisions made by students for each skill point during the consensus process of the answer data is counted as the weighted score to calculate the students' comprehensive scores. Among them, the correctness of the decision on the skill point is based on the decision results for the skill point during the teacher consensus process.

2. A teaching test method for vocational education according to claim 1, characterized in that: The user terminal is used to interact with the alliance blockchain network through an application program interface and access data on the alliance blockchain through smart contracts; The alliance blockchain is used to authenticate the user information of the user end and save the process record of students participating in the practical training project assignments through automatic execution of smart contracts, and send the answer data to the cloud platform; The cloud platform is used to build a multimodal data conversion model, convert the received student answer data into text data and score it, and then send the scoring results to the alliance blockchain.

3. A teaching test method for vocational education according to claim 2, characterized in that: The smart contract includes the following rules: User information authentication: Design an authentication mechanism to verify the identity of the user. After the authentication is passed, the student information will be stored in the alliance blockchain; Practical training project assignment release: teachers or schools serve as master nodes to release practical training project assignments. The released information includes project ID, assignment content and deadline, and is recorded on the alliance blockchain; Receiving and automatically scoring answer data: After the user submits the answer, the answer data will be sent to the alliance blockchain and marked as "pending scoring"; at the same time, the answer data will be sent to the cloud platform for automatic scoring, and the scoring results will be returned and recorded on the alliance blockchain to update the status.

4. A teaching test method for vocational education according to claim 3, characterized in that: The smart contract also includes rules: Consensus mechanism: The answer data are screened out according to the scoring results and score thresholds for consensus. The participants of the practical training project work serve as a consensus group to review each answer data and record the decision results of each skill point; New block generation: During the consensus process, a new block is generated for the answer data with the highest support. The new block contains the signatures of the supporters and the consensus result, and then the new block is broadcast to all nodes of the blockchain for data synchronization.

5. A teaching test method for vocational education according to claim 1, characterized in that: The cloud platform review and scoring comprises the following steps: Multimodal data conversion: distinguish the answer data, convert the voice data and video data to generate text descriptions, where the voice data outputs voice text through the voice recognition model, and the video data first captures the timing information of continuous frames in the video through the timing model, and then obtains the text description of the student's behavior sequence through the behavior recognition model.

6. A teaching test method for vocational education according to claim 5, characterized in that: The cloud platform review and scoring also includes the following steps: Answer data feature extraction: The converted text of the student's answer data and the reference answer are input into the BERT model for feature extraction. The text is processed through a bidirectional encoder to generate context information, and the last layer output of the BERT model is extracted as the feature vector of the answer data. Scoring model construction: Build a multi-layer perceptron network to process the feature vectors generated by BERT, introduce a gated recurrent unit model to process sequence data, capture the context information of the answer data, and fuse the output results of MLP and GRU to form the final scoring model; Obtaining answer scores: The MLP part calculates the textual implication relationship between the answer data and the reference answer to evaluate the relevance of the answer; the GRU part calculates the similarity between the answer data and the standard answer to evaluate the logic and coherence of the answer; the textual implication score and the textual similarity score are weightedly integrated to obtain the final score of the answer data.

7. A teaching test method for vocational education according to claim 1, characterized in that: The participants reached consensus on their answers, including the following steps: Select a consensus algorithm to initiate verification of the new block. Each node judges the correctness of the description of the skill point based on the answer data and gives a decision result. The decision result based on each skill point serves as a support vote for each answer data. The node collects the voting results of all participants, selects the block with the highest support according to the voting results to reach a consensus, and calculates the hash value of the block; the support includes the total number of times the consensus group supports all skill points in the answer data; The consensus block is added to the alliance blockchain and broadcast to other nodes in the network. Other nodes verify the validity of the new block and update their own ledgers.

8. A teaching test method for vocational education according to claim 1, characterized in that: The calculation formula for calculating the comprehensive score of the student is as follows: F i =S i +(W×C i ); In the formula, F i For Students i The comprehensive score, S i For Students i The basic score, W is the weight coefficient, C i For Students i The number of correct decisions.

9. A teaching test system for vocational education, used to implement a teaching test method for vocational education as claimed in any one of claims 1 to 8, characterized in that: It includes user management module, assignment publishing and answering module, answer data scoring module and skill mastery assessment module; The user management module is used to support the registration and identity authentication of students and teachers, ensure the legitimacy of users, assign permissions to different roles, manage personal information, and maintain user access to system functions; The homework publishing and answering module is used for teachers or schools to publish practical training project homework online, and students answer and submit them through user-end devices. At the same time, answer information is generated and encrypted and stored in the alliance blockchain network.

10. A teaching test system for vocational education according to claim 9, characterized in that: The answer data scoring module is used to automatically score the answer data submitted by the students and provide quick feedback through the multimodal answer data scoring model built on the cloud platform, and continuously update the scoring model; The updating of the scoring model includes the following steps: According to the teaching courses, test skill points and reference answers, add labels to the recorded practical videos or answer data, and train the multimodal conversion model through the continuously improved training data set; Collect the answer data obtained by consensus to update the training data set to train the scoring model and optimize the model structure and parameters; The skill mastery assessment module is used to assess students' mastery of the skill points in the practical training project assignment test by collecting students' answer data, reviewing the scoring results and the decision results of the consensus process on the skill points, and calculating the comprehensive score of each student.

Images