A Big Data Closed-Loop Evaluation System for the Quality of Talent Cultivation in Medical Colleges

Abstract

This invention relates to the field of talent training quality assessment technology in medical and pharmaceutical colleges, and discloses a big data closed-loop assessment system for talent training quality in medical and pharmaceutical colleges. The system includes: a data acquisition module to collect multi-source training data; a preprocessing and fusion module to construct a talent training quality dataset; a big data closed-loop assessment module containing three-level indicator construction, multi-dimensional assessment, and result analysis units to complete the assessment and generate personalized reports; an intelligent feedback module to push and visualize the assessment results; a training optimization module to adapt rigid constraints and transform them into optimization measures; and a system iteration module to ensure the assessment system is updated synchronously with the pharmaceutical industry. This invention constructs a full-link closed loop, solving problems such as assessment disconnect and low accuracy, improving the quality of talent training and its industry adaptability, and adapting to the characteristics of medical and pharmaceutical college training.

Description

Technical Field

[0001] This invention relates to the field of talent training quality assessment technology in medical colleges and universities, and specifically to a big data closed-loop assessment system for talent training quality in medical colleges and universities. Background Technology

[0002] The training of medical and pharmaceutical professionals is characterized by high specialization, strong practical application, strict entry requirements, and a long training period. The quality of this training is directly related to the development of the medical industry and public health and safety. Currently, there are significant shortcomings in the quality assessment of medical and pharmaceutical professional training: First, the closed-loop process is broken, and assessment is disconnected from training. Existing systems often focus on single aspects, such as classroom performance evaluation and internship summative assessment, lacking a comprehensive understanding of the entire process. As a result, assessment results are difficult to effectively inform adjustments to the training program, leading to a formalistic dilemma of assessment for the sake of assessment.

[0003] Secondly, data collection is fragmented and lacks comprehensive dimensions. Data sources are limited to the school's internal academic affairs system, lacking collection of core practical data. Cross-system data barriers exist, making it impossible to form a panoramic data profile of talent cultivation and supporting accurate assessment.

[0004] Third, the evaluation system suffers from weak industry adaptability and rigidity. Traditional assessment indicators largely follow general education assessment frameworks and fail to deeply integrate the characteristics of the pharmaceutical industry, such as professional qualification examination requirements, clinical job competency standards, and the evolving needs of pharmaceutical technology. Furthermore, they lack dynamic iteration mechanisms and cannot adapt to new changes in the industry.

[0005] Fourth, the accuracy of feedback and improvement is insufficient. The evaluation results are mostly macro-level conclusions, lacking detailed breakdowns of specific courses, supervising teachers, and internship units. Furthermore, adjustments to the training program are subject to rigid constraints (such as professional examination syllabi and industry standards), resulting in a time lag in feedback and improvement, and failing to achieve a secondary closed loop of evaluation, improvement, and verification.

[0006] Therefore, there is an urgent need to build a closed-loop talent training quality assessment system based on big data technology to break through existing technical bottlenecks and improve the accuracy and industry adaptability of talent training in medical colleges and universities. Summary of the Invention

[0007] The present invention aims to provide a big data closed-loop evaluation system for the quality of talent training in medical colleges and universities, in order to solve the problems of the lack of a closed loop and insufficient accuracy in the existing evaluation of the quality of talent training in medical colleges and universities.

[0008] To achieve the above objectives, the present invention adopts the following technical solution: A big data closed-loop evaluation system for the quality of talent training in medical colleges and universities includes a data acquisition module, a data preprocessing and fusion module, a big data closed-loop evaluation module, an intelligent feedback push module, a training optimization execution module, and a system iteration module. The data acquisition module is used to collect multi-source training data of students; The data preprocessing and fusion module is used to preprocess the collected multi-source training data and construct a talent training quality dataset; The big data closed-loop assessment module is used to build an assessment model for closed-loop assessment of talent training quality. The big data closed-loop assessment module includes an assessment indicator system construction unit, a multi-dimensional assessment unit, and an assessment result analysis unit. The evaluation indicator system construction unit is based on a hierarchical subdivision logic to build a three-level hierarchical evaluation indicator system. The three-level hierarchical evaluation indicator system includes the name, definition, weight, scoring criteria, and data mapping rules of each level of indicator. The multi-dimensional evaluation unit conducts static evaluation, dynamic real-time evaluation, and predictive evaluation based on a three-level hierarchical evaluation index system to obtain evaluation results. The evaluation results analysis unit breaks down the evaluation results from multiple dimensions and generates personalized evaluation reports. The intelligent feedback push module is used for pushing evaluation results and displaying them visually; The training optimization execution module is used to adapt to the rigid constraints of the training system of medical colleges and universities, and to transform feedback into specific training optimization measures; The system iteration module is used for dynamic system iteration to ensure that the evaluation system is updated in sync with the development of the pharmaceutical industry.

[0009] The principles and advantages of this solution are as follows: In practical applications, it breaks down data silos between universities, practice, and industry by collecting and standardizing multi-source heterogeneous data across all scenarios, thus solving the problems of fragmented and incomplete dimensions in traditional assessment data; it addresses the issues of vague assessment standards, lack of industry adaptability, and the inability to achieve only a single dimension assessment through a three-level hierarchical assessment indicator system and a multi-dimensional assessment model; it ensures accurate delivery of assessment results through intelligent feedback push, solving the problems of inaccurate feedback and insufficient collaboration among various stakeholders; it adapts the training optimization execution module to the rigid constraints of medical colleges, solving the problems of assessment results failing to be implemented as training measures and optimization becoming a mere formality; it enables dynamic upgrades through a system iteration module, solving the problems of a rigid assessment system and disconnection from industry development; and ultimately, it constructs a closed-loop system encompassing collection, processing, assessment, feedback, optimization, and iteration, resolving the problems of existing assessment loop breaks and disconnection from training, thereby improving the quality of medical talent training and its suitability for industry positions.

[0010] Preferably, as an improvement, the data acquisition module includes an on-campus core data acquisition unit, a practical scenario data acquisition unit, an industry feedback data acquisition unit, and a dynamically supplementary data acquisition unit; The core data collection unit within the school connects to the school's data system via API interfaces to collect students' basic information, theoretical teaching data, and data on ideological and political education and medical ethics performance. The practical scenario data acquisition unit is used to build a collaborative data acquisition channel between schools and enterprises and schools and colleges to collect practical training data; The industry feedback data collection unit is used to connect with pharmaceutical industry associations, employers, and professional qualification examination institutions to collect data on graduates' professional qualification examination pass rates, job competency scores, employer satisfaction, updated industry job demand information, and graduates' career development data. The dynamic data acquisition unit uploads real-time dynamic data via a mobile app.

[0011] Technical benefits: It facilitates the comprehensiveness of data collection and solves the problems of single data source and poor timeliness in traditional assessments.

[0012] Preferably, as an improvement, the evaluation index system construction unit uses the analytic hierarchy process to assign weights to the third-level indicators, maps the data in the talent training quality dataset to the third-level indicators, and quantifies the scores from 1 to 100, calculating the scores level by level according to the weights; The primary indicators include the achievement of training objectives, process quality, outcome quality, and industry suitability; the secondary indicators include course quality, internship quality, professional competence, and job satisfaction; and the tertiary indicators include prescription review accuracy, medical record writing quality, and professional examination pass rate.

[0013] Technical benefits: It solves the problems of vague and disorganized traditional evaluation indicators, and improves the standardization and professionalism of the evaluation.

[0014] Preferably, as an improvement, the static evaluation includes using the analytic hierarchy process (AHP) based on completed training cycle data to calculate historical scores and comprehensive scores for each level of indicators, comparing the levels of similar majors across schools and different cohorts within the same school, and outputting a static evaluation report. The static evaluation report includes historical training quality scores, inter-school / cohort comparison results, and training objective achievement analysis. The dynamic real-time evaluation includes collecting culture process data in real time based on the current culture cycle, using the random forest algorithm to analyze the correlation between real-time process data and tertiary indicators, dynamically calculating process scores, comparing with static evaluation benchmarks, identifying deviations and issuing warnings, and outputting a dynamic evaluation warning report. The dynamic evaluation warning report includes real-time process quality scores, deviation point locations, and immediate improvement suggestions. The prediction and evaluation uses an LSTM neural network to train a prediction model based on static and dynamic evaluation results, predicts the changing trends of evaluation indicators, and outputs a prediction and evaluation report. The prediction and evaluation report includes indicator trend curves, prediction scores, risk point prompts, and forward-looking optimization directions.

[0015] Technical benefits: Static, dynamic, and predictive assessments work in synergy, covering historical, real-time, and future time dimensions, solving the problem that traditional assessments can only achieve assessments in a single time dimension and cannot predict risks.

[0016] Preferably, as an improvement, the intelligent feedback push module includes: Personalized feedback units push customized feedback information based on the different needs of different users; The visualization unit uses BI visualization tools to build an evaluation result display platform, displaying evaluation indicators in the form of line charts, bar charts, and heat maps, and supporting data drill-down queries; The feedback and interaction unit establishes a feedback and interaction channel, supporting users to appeal against or supplement their opinions on the evaluation results.

[0017] Technical benefits: It facilitates adaptation to the needs of different stakeholders, solves the problems of traditional feedback being one-size-fits-all and lacking specificity, and ensures that each stakeholder obtains accurate and usable assessment information.

[0018] Preferably, as an improvement, the cultivation optimization execution module includes: The curriculum optimization unit intelligently generates suggestions for adjusting the curriculum based on evaluation results and industry needs, and connects with the university's curriculum management system. The teaching process optimization unit is used to push out teaching optimization tasks and monitor the implementation process of optimization measures in real time. The practical teaching optimization unit is used to push teaching optimization tasks to the practical end; The personalized training optimization unit generates personalized training plans based on students' skill gaps, which are then pushed to students via a mobile app to track the progress of the plans.

[0019] Technical effects: It optimizes all aspects of the program, teaching, practice and individuals, adapts to the rigid constraints of the medical school training system, and solves the problems of traditional evaluation results not being able to be implemented and optimization measures being fragmented.

[0020] Preferably, as an improvement, the system iteration module includes: The indicator iteration unit regularly updates the evaluation indicator system based on changes in industry needs, feedback on evaluation results, and expert opinions. The algorithm iteration unit periodically iterates and optimizes the evaluation model of the multi-dimensional evaluation unit.

[0021] Technical benefits: It ensures the long-term adaptability and advancement of the evaluation system, prevents the system from becoming outdated due to industry development, and supports the continuous improvement of talent training quality.

[0022] Preferably, as an improvement, the blockchain technology in the data preprocessing and fusion module adopts a consortium blockchain architecture, with nodes including university nodes, practice base nodes, industry association nodes, and regulatory nodes, and the consensus mechanism of each node adopts the PBFT algorithm.

[0023] Technical benefits: Facilitates the security and efficiency of cross-domain data fusion. Attached Figure Description

[0024] Figure 1 This is a schematic diagram of a big data closed-loop evaluation system for the quality of talent training in medical colleges. Detailed Implementation

[0025] The following detailed description illustrates the specific implementation method: The basic implementation examples are as follows: Figure 1 As shown, a big data closed-loop evaluation system for the quality of talent cultivation in medical and pharmaceutical colleges includes a data acquisition module, a data preprocessing and fusion module, a big data closed-loop evaluation module, an intelligent feedback and push module, a training optimization execution module, and a system iteration module. These modules work together to achieve a closed-loop operation of data acquisition, multi-dimensional evaluation, precise feedback, dynamic optimization, and iterative upgrades. Specifically, the system constructs a medical and pharmaceutical talent cultivation quality dataset through the full-scenario acquisition and standardized fusion of multi-source heterogeneous data. Based on a hierarchical evaluation model, static, dynamic, and predictive evaluations are completed. The evaluation results are accurately pushed to various stakeholders and transformed into implementable training optimization measures. Finally, the system iterates itself based on industry needs and operational results, ensuring dynamic adaptation of the evaluation system to the characteristics of medical and pharmaceutical talent cultivation and the needs of industry development.

[0026] The data acquisition module is used to collect multi-source student training data. Specifically, the data acquisition module includes an internal core data acquisition unit, a practical scenario data acquisition unit, an industry feedback data acquisition unit, and a dynamically supplementary data acquisition unit. The internal core data acquisition unit connects with the internal data system via API to collect basic student information, theoretical teaching data, and data on ideological and political education and medical ethics performance. The internal data system includes the university's academic affairs management system, student affairs system, experimental teaching management system, and scientific research management system.

[0027] The practical scenario data collection unit is used to build a collaborative data collection channel between schools and enterprises, and schools and colleges to collect practical training data. The practical training data includes clinical internship data, pharmacy practical training data, and public health practical training data. Among them, the clinical internship data includes department rotation records, number of consultations, medical record writing quality scores, surgical cooperation proficiency, real-time comments from teaching staff, and doctor-patient communication case records.

[0028] The industry feedback data collection unit is used to connect with pharmaceutical industry associations, employers, and professional qualification examination institutions to collect data on graduates' professional qualification examination pass rates, job competency scores, employer satisfaction, updated industry job requirements, and graduates' career development.

[0029] The dynamic data collection unit uploads real-time dynamic data via a mobile app. The mobile app supports students, teachers, and mentors in uploading real-time dynamic data in text, image, video, and audio formats, including internship logs, teaching reflections, feedback on mentoring issues, and students' career planning requests.

[0030] The data preprocessing and fusion module is used to preprocess the collected multi-source training data to construct a talent training quality dataset. Preprocessing includes cleaning, standardization, and fusion; an outlier detection algorithm (a Z-score algorithm optimized based on the characteristics of pharmaceutical industry data) is used to remove invalid data (such as blank internship logs and incorrect grade entry), correct missing data (through interpolation of data from adjacent periods and comparison with data from similar students), and filter duplicate data. A pharmaceutical professional data standard system is established, uniformly encoding data from different sources and in different formats (such as course codes, internship position codes, and competency indicator codes). Qualitative data (such as teaching comments and medical ethics performance) is converted into quantitative scores (1-10 points) through semantic analysis, enabling data comparison and computation. A data fusion channel is built based on blockchain technology to ensure data security and immutability (adapting to the privacy data of pharmaceutical talents and the business information protection needs of employers). A federated learning algorithm is used to achieve cross-domain fusion of on-campus data, practical data, and industry data, generating a full-cycle training data profile for each student, a training quality dataset for each major, and a teaching quality dataset for each practical base. It adopts a hybrid storage architecture of "distributed database + time-series database". The distributed database stores static data (basic information, academic records) and the time-series database stores dynamic data (internship process data, real-time feedback data), supporting efficient storage and fast query of massive data.

[0031] The blockchain technology in the data preprocessing and fusion module adopts a consortium blockchain architecture, with nodes including university nodes, practice base nodes, industry association nodes, and regulatory nodes. The consensus mechanism of each node adopts the PBFT algorithm.

[0032] The big data closed-loop assessment module is used to build assessment models for closed-loop assessment of talent cultivation quality. The module includes an assessment indicator system construction unit, a multi-dimensional assessment unit, and an assessment result analysis unit.

[0033] The evaluation index system construction unit is based on the logic of hierarchical subdivision to construct a three-level hierarchical evaluation index system. The three-level hierarchical evaluation index system includes the name, definition, weight, scoring standard and data mapping rules of each level of index. The first-level index includes the achievement of training objectives, process quality, result quality and industry adaptability. The second-level index includes course quality, internship quality, professional competence and job satisfaction. The third-level index includes prescription review accuracy, medical record writing quality and professional examination pass rate. The division of the three-level index system follows the logic of "general-specific-detailed", which is in line with the needs of medical colleges and universities for talent training of "goal orientation, process control, result implementation and industry adaptability". At the same time, it connects the professional standards and job competency requirements of the medical industry. The specific basis is as follows: (1) The first-level index anchors the core dimension of talent training quality. Around the key nodes of the entire cycle of medical talent training, it covers the entire chain of goal setting, process implementation, result output and industry inspection, solves the problem of fragmentation of traditional evaluation dimensions, and ensures the comprehensiveness of evaluation. Among them, industry adaptability is a core index exclusive to medical professions, which is adapted to the characteristics of strong professional access and high job adaptability requirements of medical talents. (2) Secondary indicators: Decompose the primary indicators into feasible assessment directions. The four primary indicators are decomposed into specific assessment scenarios, taking into account both on-campus training and industry verification: the achievement of training objectives and process quality are decomposed into core on-campus scenarios (course quality, internship quality), and the quality of results and industry adaptability are decomposed into industry-related scenarios (professional competence, job satisfaction), so as to achieve the connection between on-campus assessment and industry assessment and avoid the disconnect between assessment and job requirements. (3) Tertiary indicators: Refine into specific indicators that can be quantified and measured. Focusing on the core competencies and professional requirements of the pharmaceutical profession, indicators such as prescription review accuracy rate, medical record writing quality, and professional examination pass rate are selected, which can be directly collected and quantified to solve the problem that traditional assessment is mainly qualitative and difficult to implement. All tertiary indicators are aligned with pharmaceutical industry standards (such as medical record writing conforming to clinical diagnosis and treatment guidelines, prescription review conforming to pharmaceutical management standards, and professional examination pass rate conforming to professional access requirements), to ensure the professionalism and practicality of the assessment.

[0034] The tertiary indicators, as the smallest quantitative unit of evaluation, provide layer-by-layer support to the secondary and primary indicators through weight allocation and data mapping, ultimately forming a comprehensive evaluation result. The specific application process is as follows: (1) Weight allocation of indicators: The analytic hierarchy process is adopted, and the opinions of experts are combined to allocate weights to indicators at all levels (e.g., the weight of "industry suitability" in the first-level indicators is no less than 25%, the weight of "internship quality" in the second-level indicators is no less than 30%, and the weight of "professional examination pass rate" in the third-level indicators is no less than 20%), highlighting the characteristics of practice orientation and industry orientation of the medical profession.

[0035] (2) Data mapping and scoring: The multi-source data (such as prescription review data, medical records, and professional examination results) obtained by the data acquisition module are mapped with the corresponding three-level indicators and converted into quantitative scores (1-100 points) through standardized formulas; then, according to the weights, the scores of the second-level indicators and the scores of the first-level indicators are calculated and summarized at each level to finally form the comprehensive score of talent training quality.

[0036] (3) Dynamic adaptation and adjustment: The three-level indicators are dynamically added or optimized according to changes in pharmaceutical industry policies (such as iteration of treatment guidelines and updates to the professional examination syllabus) and adjustments in job requirements (such as new requirements for clinical skills in smart healthcare) to ensure that the evaluation criteria are in sync with industry development.

[0037] The multi-dimensional assessment unit utilizes a three-tiered hierarchical assessment indicator system to conduct static, dynamic, real-time, and predictive assessments, yielding evaluation results. These three assessments complement each other across historical, real-time, and future time dimensions, covering the long-cycle and slow-feedback characteristics of pharmaceutical talent development and avoiding the limitations of a single assessment dimension.

[0038] The static evaluation includes using the analytic hierarchy process (AHP) based on completed training cycle data to calculate historical scores and comprehensive scores for each level of indicators, comparing the level of similar majors across universities and the level of different cohorts within the university, and outputting a static evaluation report. The static evaluation report includes historical training quality scores, inter-university / cohort comparison results, and analysis of the achievement of training objectives.

[0039] The dynamic real-time evaluation includes collecting culture process data in real time based on the current culture cycle, using the random forest algorithm to analyze the correlation between real-time process data and tertiary indicators, dynamically calculating process scores, comparing with static evaluation benchmarks, identifying deviations and issuing warnings, and the dynamic evaluation warning report includes real-time process quality scores, deviation point locations, and immediate improvement suggestions.

[0040] The prediction and evaluation uses an LSTM neural network to train a prediction model based on static and dynamic evaluation results. It predicts the changing trends of evaluation indicators (such as the pass rate of professional examinations for subsequent graduates and job suitability) and outputs a prediction and evaluation report. The prediction and evaluation report includes indicator trend curves, prediction scores, risk point tips, and forward-looking optimization directions.

[0041] The assessment results analysis unit is used to break down assessment results from multiple dimensions and generate personalized assessment reports. For example, assessment results can be broken down by subject and scenario to form tiered reports: individual student reports (shortcomings and improvement plans), professional level reports (core indicator performance and weaknesses), practice base reports (teaching quality and suitability), and school-level reports (overall training quality and industry fit gaps). The analysis conclusions and personalized reports are then pushed to the intelligent feedback module to provide a basis for subsequent training optimization.

[0042] The intelligent feedback push module is used for pushing and visualizing evaluation results. This module includes a personalized feedback unit, a visualization unit, and a feedback interaction unit. The personalized feedback unit pushes customized feedback information based on the different needs of different users, such as students, teachers, supervising teachers, teaching management departments, and employers.

[0043] The visualization unit uses BI visualization tools to build an evaluation result display platform, displaying evaluation indicators in the form of line charts, bar charts, and heat maps, supporting data drill-down queries, and making the evaluation results intuitive and traceable.

[0044] The feedback and interaction unit establishes a feedback and interaction channel, supporting users to appeal against or supplement their opinions on the evaluation results.

[0045] The training optimization execution module is designed to adapt to the rigid constraints of the medical and pharmaceutical colleges' training systems, transforming feedback into specific training optimization measures. This module includes a training program optimization unit, a teaching process optimization unit, a practical teaching optimization unit, and a personalized training optimization unit. The training program optimization unit intelligently generates training program adjustment suggestions based on evaluation results and industry needs, connecting to the college's training program management system. The teaching process optimization unit pushes teaching optimization tasks and monitors the implementation process of optimization measures in real time. The practical teaching optimization unit pushes teaching optimization tasks to the practical training end. The personalized training optimization unit generates personalized training plans targeting students' skill gaps, pushes them to students via a mobile app, and tracks the plan's progress.

[0046] The system iteration module is used for dynamic system iteration, ensuring that the evaluation system is updated in sync with the development of the pharmaceutical industry. This module includes an indicator iteration unit and an algorithm iteration unit. The indicator iteration unit regularly updates the evaluation indicator system based on changes in industry needs, evaluation result feedback, and expert opinions. The algorithm iteration unit periodically iterates and optimizes the evaluation model of the multi-dimensional evaluation units, adjusting the LSTM neural network weight matrix and bias terms, and optimizing the weight calculation method of the analytic hierarchy process.

[0047] The above descriptions are merely embodiments of the present invention, and common knowledge such as specific technical solutions and / or characteristics are not described in detail here. It should be noted that those skilled in the art can make various modifications and improvements without departing from the technical solutions of the present invention, and these should also be considered within the scope of protection of the present invention. These modifications and improvements will not affect the effectiveness of the implementation of the present invention or the practicality of the patent. The scope of protection claimed in this application should be determined by the content of its claims, and the specific embodiments described in the specification can be used to interpret the content of the claims.

Claims

1. A big data closed-loop evaluation system for the quality of talent cultivation in medical colleges, characterized in that, It includes a data acquisition module, a data preprocessing and fusion module, a big data closed-loop evaluation module, an intelligent feedback push module, a training optimization execution module, and a system iteration module; The data acquisition module is used to collect multi-source training data of students; The data preprocessing and fusion module is used to preprocess the collected multi-source training data and construct a talent training quality dataset; The big data closed-loop assessment module is used to build an assessment model for closed-loop assessment of talent training quality. The big data closed-loop assessment module includes an assessment indicator system construction unit, a multi-dimensional assessment unit, and an assessment result analysis unit. The evaluation indicator system construction unit is based on a hierarchical subdivision logic to build a three-level hierarchical evaluation indicator system. The three-level hierarchical evaluation indicator system includes the name, definition, weight, scoring criteria, and data mapping rules of each level of indicator. The multi-dimensional evaluation unit conducts static evaluation, dynamic real-time evaluation, and predictive evaluation based on a three-level hierarchical evaluation index system to obtain evaluation results. The evaluation results analysis unit breaks down the evaluation results from multiple dimensions and generates personalized evaluation reports. The intelligent feedback push module is used for pushing evaluation results and displaying them visually; The training optimization execution module is used to adapt to the rigid constraints of the training system of medical colleges and universities, and to transform feedback into specific training optimization measures; The system iteration module is used for dynamic system iteration to ensure that the evaluation system is updated in sync with the development of the pharmaceutical industry.

2. The big data closed-loop evaluation system for the quality of talent cultivation in medical colleges and universities according to claim 1, characterized in that: The data acquisition module includes a core data acquisition unit within the school, a practical scenario data acquisition unit, an industry feedback data acquisition unit, and a dynamically supplementary data acquisition unit. The core data collection unit within the school connects to the school's data system via API interfaces to collect students' basic information, theoretical teaching data, and data on ideological and political education and medical ethics performance. The practical scenario data acquisition unit is used to build a collaborative data acquisition channel between schools and enterprises and schools and colleges to collect practical training data; The industry feedback data collection unit is used to connect with pharmaceutical industry associations, employers, and professional qualification examination institutions to collect data on graduates' professional qualification examination pass rates, job competency scores, employer satisfaction, updated industry job demand information, and graduates' career development data. The dynamic data acquisition unit uploads real-time dynamic data via a mobile app.

3. The big data closed-loop evaluation system for the quality of talent training in medical colleges and universities according to claim 1, characterized in that: The evaluation index system construction unit uses the analytic hierarchy process to assign weights to the three-level indicators, maps the data in the talent training quality dataset to the three-level indicators, and quantifies the scores from 1 to 100, calculating the scores level by level according to the weights. The primary indicators include the achievement of training objectives, process quality, outcome quality, and industry suitability; the secondary indicators include course quality, internship quality, professional competence, and job satisfaction; and the tertiary indicators include prescription review accuracy, medical record writing quality, and professional examination pass rate.

4. The big data closed-loop evaluation system for the quality of talent training in medical colleges and universities according to claim 1, characterized in that: The static evaluation includes using the analytic hierarchy process (AHP) based on completed training cycle data to calculate historical scores and comprehensive scores for each level of indicators, comparing the level of similar majors across schools and the level of different cohorts within the school, and outputting a static evaluation report. The static evaluation report includes historical training quality scores, inter-school / cohort comparison results, and analysis of the achievement of training objectives. The dynamic real-time evaluation includes collecting culture process data in real time based on the current culture cycle, using the random forest algorithm to analyze the correlation between real-time process data and tertiary indicators, dynamically calculating process scores, comparing with static evaluation benchmarks, identifying deviations and issuing warnings, and outputting a dynamic evaluation warning report. The dynamic evaluation warning report includes real-time process quality scores, deviation point locations, and immediate improvement suggestions. The prediction and evaluation uses an LSTM neural network to train a prediction model based on static and dynamic evaluation results, predicts the changing trends of evaluation indicators, and outputs a prediction and evaluation report. The prediction and evaluation report includes indicator trend curves, prediction scores, risk point prompts, and forward-looking optimization directions.

5. A big data closed-loop evaluation system for the quality of talent cultivation in medical colleges and universities according to claim 1, characterized in that, The intelligent feedback push module includes: Personalized feedback units push customized feedback information based on the different needs of different users; The visualization unit uses BI visualization tools to build an evaluation result display platform, displaying evaluation indicators in the form of line charts, bar charts, and heat maps, and supporting data drill-down queries; The feedback and interaction unit establishes a feedback and interaction channel, supporting users to appeal against or supplement their opinions on the evaluation results.

6. A big data closed-loop evaluation system for the quality of talent cultivation in medical colleges and universities according to claim 1, characterized in that, The cultivation optimization execution module includes: The curriculum optimization unit intelligently generates suggestions for adjusting the curriculum based on evaluation results and industry needs, and connects with the university's curriculum management system. The teaching process optimization unit is used to push out teaching optimization tasks and monitor the implementation process of optimization measures in real time. The practical teaching optimization unit is used to push teaching optimization tasks to the practical end; The personalized training optimization unit generates personalized training plans based on students' skill gaps, which are then pushed to students via a mobile app to track the progress of the plans.

7. A big data closed-loop evaluation system for the quality of talent cultivation in medical colleges and universities according to claim 4, characterized in that, The system iteration module includes: The indicator iteration unit regularly updates the evaluation indicator system based on changes in industry needs, feedback on evaluation results, and expert opinions. The algorithm iteration unit periodically iterates and optimizes the evaluation model of the multi-dimensional evaluation unit.

8. A big data closed-loop evaluation system for the quality of talent cultivation in medical colleges and universities according to claim 1, characterized in that: The blockchain technology in the data preprocessing and fusion module adopts a consortium blockchain architecture, with nodes including university nodes, practice base nodes, industry association nodes, and regulatory nodes. The consensus mechanism of each node adopts the PBFT algorithm.

Images