A dynamic multi-dimensional student ability evaluation and career planning method and system

Abstract

The application relates to a dynamic multi-dimensional student ability evaluation and career planning method and system, which comprises the following steps: collecting structured and unstructured student ability data from an education information system, integrating the data to form an original data set, calling a standardization algorithm to generate a feature vector set containing multi-dimensional ability indexes, obtaining market demand data through an external interface, combining a preset professional ability matrix rule, inputting the feature vector set into a multi-dimensional ability evaluation model, and outputting an ability evaluation result set, calling a visualization engine based on the result set to generate an ability distribution radar chart and an analysis report, and outputting a visualization chart and the report, receiving a user development direction selection instruction, applying a path planning algorithm to generate personalized planning suggestions and progressive improvement paths, and outputting a planning report, obtaining new ability data based on the planning report, calling a deviation detection model to monitor the improvement progress, and triggering an early warning when the threshold is exceeded. The application improves the scientificity and effectiveness of student career development guidance.

Description

Technical Field

[0001] This application relates to the field of educational informatization technology, and in particular to a dynamic, multi-dimensional student ability assessment and career planning method and system. Background Technology

[0002] With the deep integration of information technology into education and the comprehensive advancement of the digital education strategy, the cultivation of students' comprehensive qualities and personalized career development planning have become core concerns of higher education reform. Traditional academic affairs management, student affairs systems, and various online learning platforms have accumulated a large amount of structured data (such as course grades and attendance records) and unstructured data (such as practice reports and teacher comments) related to student growth. Meanwhile, data intelligence technologies, represented by big data analysis and machine learning, offer new methods for deeply exploring students' abilities and predicting their development potential. Against this backdrop, using data-driven methods to assist students in conducting scientific and systematic self-awareness and career planning has become an important direction for educational technology development, leading to the emergence of various student ability assessment tools and career planning platforms.

[0003] However, in practice, most assessment systems rely on fixed indicators and weights, and their assessment standards are often based on historical experience or general models. This makes it difficult to respond sensitively to rapid changes in labor market demands and dynamic shifts in industry capabilities, resulting in assessment results that lag behind and are disconnected from students' future professional environments. Secondly, the planning suggestions provided by existing systems are often too general, failing to be based on detailed gap analysis and lacking in-depth, dynamic mapping and sequencing with rich educational resources such as courses and practical training, leading to insufficient operability of the planning paths. Furthermore, the process from planning to goal achievement is long-term; many plans become mere formalities, unable to be dynamically adjusted according to students' actual progress, reducing the effectiveness of career development guidance and the rate of goal achievement. Summary of the Invention

[0004] To address the aforementioned technical issues, this application provides a dynamic, multi-dimensional student ability assessment and career planning method and system.

[0005] Firstly, this application provides a dynamic, multi-dimensional student ability assessment and career planning method, employing the following technical solution: Structured and unstructured student ability data are collected from the education information system, and the raw dataset is output after data integration. The original dataset is cleaned and quantized using a standardization algorithm to generate feature vectors containing multi-dimensional capability indicators, and a standardized feature vector set is output. Market demand data is obtained through an external data interface. Combined with the preset professional competence matrix rules and the market demand data, the standardized feature vector set is input into a pre-constructed multi-dimensional competence assessment model for rule matching and weight coefficient update. The competence assessment score is calculated and the competence assessment result set is output. Based on the capability assessment results set, the visualization engine is invoked to generate a capability distribution radar chart and automatically generate an analysis report, outputting visualization charts and analysis reports; It receives development direction selection instructions from the user interface, applies path planning algorithms to compare ability differences and map resources, generates personalized career planning suggestions and progressive ability improvement paths, and outputs a personalized planning report. Based on the personalized planning report, newly collected student ability data is obtained, and a deviation detection model is called to monitor the deviation in the progress of student ability improvement. When the deviation exceeds the threshold, an early warning notification is triggered.

[0006] By adopting the above technical solutions, a highly intelligent and adaptive student career planning method has been achieved. Its technical effects are reflected in three aspects: First, the dynamic and forward-looking nature of the assessment standards. By connecting with market data and adjusting model weights in real time, the assessment results closely align with changes in social needs, overcoming the lag of traditional assessments. Second, the accuracy and feasibility of the planning path. Based on quantified ability gaps, resource mapping and sequential planning are conducted, ensuring that the advice students receive is no longer vague but specific, orderly, and personalized action plans. Third, the closed-loop nature and timely intervention of the management process. Through continuous deviation monitoring and early warning, career planning is transformed from a static "one-off suggestion" into a dynamic "full-process growth support system," effectively improving the implementation effect of the plan and the sustainability of student development.

[0007] Secondly, this application provides a dynamic, multi-dimensional student ability assessment and career planning system, which adopts the following technical solution: The multi-source data integration and consolidation module is used to collect structured and unstructured student ability data from the education information system, integrate the data, and output the original dataset. The standardized vector generation module is used to clean and quantize the original dataset by calling a standardization algorithm, generate feature vectors containing multi-dimensional capability indicators, and output a standardized feature vector set. The dynamic capability assessment module is used to obtain market demand data through an external data interface, combine the preset professional capability matrix rules and the market demand data, input the standardized feature vector set into a pre-built multi-dimensional capability assessment model for rule matching and weight coefficient update, calculate the capability assessment score and output the capability assessment result set. The visualization analysis module is used to generate a capability distribution radar chart based on the capability assessment result set, and automatically generate an analysis report, outputting visualization charts and analysis reports. The personalized path planning module receives development direction selection instructions from the user interface, applies path planning algorithms to compare ability differences and map resources, generates personalized career planning suggestions and progressive ability improvement paths, and outputs a personalized planning report. The progress monitoring and feedback module is used to acquire newly collected student ability data based on personalized planning reports, call the deviation detection model to monitor deviations in student ability improvement progress, and trigger an early warning notification when the deviation exceeds a threshold.

[0008] Thirdly, this application provides a computer-readable storage medium, which adopts the following technical solution: A computer-readable storage medium storing a computer program that can be loaded by a processor and executed as in any of the methods in the first aspect.

[0009] In summary, this application includes at least one of the following beneficial technical effects: by dynamically integrating multi-source student ability data, standardizing and quantifying the data, and combining it with a multi-dimensional evaluation model based on market demand, it achieves accurate profiling and dynamic tracking of student abilities; relying on visualization analysis and personalized path planning, it provides students with progressive improvement plans tailored to their strengths and development goals; and by using deviation monitoring and early warning mechanisms to intervene in the growth process in real time, it effectively bridges the gap between ability development and career needs. This not only helps students clarify their career direction and enhance their core competitiveness, but also provides data support for educational institutions to optimize their training strategies and for enterprises to accurately match talent, promoting the dynamic adaptation of talent cultivation and the employment market, and improving the scientific nature and effectiveness of career development guidance. Attached Figure Description

[0010] Figure 1 This is a first flowchart illustrating a dynamic multi-dimensional student ability assessment and career planning method according to one embodiment of this application.

[0011] Figure 2 This is a second flowchart illustrating a dynamic multi-dimensional student ability assessment and career planning method according to one embodiment of this application.

[0012] Figure 3 This is a schematic diagram of the third process of a dynamic multi-dimensional student ability assessment and career planning method according to one embodiment of this application.

[0013] Figure 4 This is a schematic diagram of the fourth process of a dynamic multi-dimensional student ability assessment and career planning method according to one embodiment of this application.

[0014] Figure 5This is a schematic diagram of the fifth process of a dynamic multi-dimensional student ability assessment and career planning method according to one embodiment of this application.

[0015] Figure 6 This is a schematic diagram of the sixth process of a dynamic multi-dimensional student ability assessment and career planning method according to one embodiment of this application.

[0016] Figure 7 This is a schematic diagram of the seventh process of a dynamic multi-dimensional student ability assessment and career planning method according to one embodiment of this application.

[0017] Figure 8 This is a schematic diagram of the eighth process of a dynamic multi-dimensional student ability assessment and career planning method according to one embodiment of this application. Detailed Implementation

[0018] To make the purpose, technical solution, and advantages of this application clearer, the following description is provided in conjunction with the appendix. Figures 1-8 The present application will be further described in detail below with reference to embodiments. It should be understood that the specific embodiments described herein are for illustrative purposes only and are not intended to limit the scope of the application.

[0019] This application discloses a dynamic, multi-dimensional student ability assessment and career planning method.

[0020] Reference Figure 1 A dynamic, multi-dimensional student ability assessment and career planning method, specifically including: Step S101: Collect structured and unstructured student ability data from the education information system, integrate the data, and output the original dataset. The educational information system is a general term for a collection of data sources, specifically including the academic affairs management system (which mainly stores structured data such as course grades and credits), the student information database (which may contain unstructured records such as student status, awards and punishments), and the online assessment platform (which generates unstructured results such as psychological assessments and career aptitude tests).

[0021] Specifically, structured data refers to information stored in database tables with fixed formats and fields (such as "student ID, course, and score" in a grade table); unstructured data refers to information with free formats, such as text reports, comments, and project documents, which require parsing to extract value (such as descriptions in practice records and teacher comments).

[0022] In this embodiment of the application, data integration mainly involves establishing a unified student identity identifier (such as student ID), and associating, aligning and initially aggregating data from different sources and in different formats according to timeline and event type, to form an original dataset centered on individual students and containing multi-dimensional behavioral and achievement records.

[0023] Step S102: Clean and quantize the original dataset using a standardization algorithm to generate feature vectors containing multi-dimensional capability indicators and output a standardized feature vector set. The original dataset contains noise (such as data entry errors), missing values, and incomparability (such as percentage scores and graded evaluations). Data cleaning can ensure data quality through rules (such as removing values ​​that are outside a reasonable range) or algorithms (such as filling missing values ​​with the mean / median). Quantitative processing aims to concretize abstract capabilities into numerical indicators.

[0024] In this embodiment, the system pre-defines five core dimensions, including academic ability and innovation ability, and designs sub-indicators for each dimension. For example, "innovation ability" can be further refined into "project originality" and "solution complexity," etc. These indicators need to obtain initial scores by analyzing project reports through natural language processing (NLP) or by mapping competition award levels through rules. The ultimate function of standardization algorithms (such as Min-Max normalization and Z-Score standardization) is to transform the scores of all indicators to a unified scale (such as the 0-1 range), eliminating biases caused by different scoring standards.

[0025] Ultimately, each student is represented as an eigenvector in mathematical terms, which can be viewed as a point in a multidimensional space. Each dimension of the vector corresponds to a standardized ability score. The eigenvectors of all students constitute a standardized eigenvector set, which is the data foundation for scientific assessment and comparison.

[0026] Step S103: Obtain market demand data through an external data interface, combine the preset professional competence matrix rules and market demand data, input the standardized feature vector set into the pre-built multi-dimensional competence assessment model for rule matching and weight coefficient update, calculate the competence assessment score and output the competence assessment result set; The underlying logic of this step is to move the evaluation criteria away from static, subjective settings and align them in real-time with the dynamic demands of the external labor market and the internal professional training objectives. Market demand data can be obtained by periodically crawling or accessing data from recruitment websites, industry analysis reports, and company surveys through external data interfaces. The professional competency matrix rules are a knowledge framework pre-defined by education experts to describe the core competencies that graduates of a specific major should possess and their relative importance.

[0027] In this embodiment, rule matching and weight coefficient updating is a dynamic decision-making process: the system analyzes market demand data, identifies "emerging capability demand tags" (such as AI-powered operations and maintenance, digital marketing), and then performs semantic matching with existing capability items in the professional capability matrix. If highly relevant emerging demands are found, the matrix can be expanded or the weights of existing capabilities can be adjusted. For example, the weight of "data analysis" in a traditional financial professional matrix may be increased due to fintech trends. Next, the multi-dimensional capability assessment model (which covers five dimensions and specific assessment methods) calculates a weighted sum of each student's standardized feature vectors based on the updated weight coefficients.

[0028] Specifically, the model multiplies the score of each dimension of the feature vector by its dynamically adjusted weight, and then sums them to obtain the final total ability assessment score and scores for each dimension, forming an ability assessment result set. This makes the student's ability score no longer a fixed value, but a timely and dynamic evaluation that reflects current market and professional expectations.

[0029] Step S104: Based on the capability assessment result set, call the visualization engine to generate a capability distribution radar chart and automatically generate an analysis report, outputting visualization charts and analysis reports; Specifically, the visualization engine (such as ECharts or D3.js) receives the ability assessment results set. Its key technical action is to draw a radar chart (also known as a spider chart) in a polar coordinate system based on the scores of each dimension. The polygonal shape of the radar chart can very intuitively show the "outline" and "balance" of the student's ability structure: an outward-protruding vertex represents a strength ability, while an inward-concave vertex represents a weakness that needs to be improved.

[0030] The design uses solid lines to represent the current level and dashed lines to represent the target level, creating a clear visual contrast that allows students to instantly identify their shortcomings. Meanwhile, the automatic report generation function relies on preset text generation templates and a rule engine. Based on score levels, comparisons between dimensions, and the gap with the target line, the system automatically organizes language to describe strengths, point out weaknesses, and fills in relevant typical interpretations and suggestions from the knowledge base. This combination of visual charts and text reports transforms the abstract vector of scores into a concrete and actionable self-awareness.

[0031] Step S105: Receive the development direction selection instruction input from the user interface, apply the path planning algorithm to compare ability differences and map resources, generate personalized career planning suggestions and progressive ability improvement paths, and output a personalized planning report. The core logic of this step is optimal path planning under resource constraints. Students select directions such as "further education" or "employment" through the UI, which sets the target node for path planning.

[0032] Subsequently, the path planning algorithm begins its work: First, it compares the ability gaps, calculating the difference between the student's current ability feature vector and the ability model vector required by the target direction, accurately identifying the specific dimensions and extent that need improvement. Next is resource mapping. The algorithm searches a knowledge graph containing resources such as courses, internships, training, certifications, and competitions for resource items that can effectively bridge the identified ability gaps, and evaluates their required time and sequential dependencies.

[0033] Ultimately, the algorithm generates a progressive skill enhancement path, typically a timeline plan divided into foundational, reinforcement, and goal-oriented phases. Each phase is configured with a specific combination of recommended resources (e.g., "Take the 'Python Data Analysis' course this semester and participate in a data analysis internship at a company this summer"). Essentially, this path deconstructs macro-level career goals into a series of ordered, achievable, and personalized micro-learning and action tasks, forming an executable, personalized planning report.

[0034] Step S106: Based on the personalized planning report, acquire newly collected student ability data, call the deviation detection model to monitor the deviation of student ability improvement progress, and trigger an early warning notification when the deviation exceeds the threshold.

[0035] The system uses the phased ability improvement goals set in the personalized planning report as the expected baseline. As time goes by, the system continuously acquires the latest activity and performance data of students, and after processing, obtains new ability feature vectors.

[0036] In this embodiment, the core logic of the deviation detection model is to calculate the rate of change and compare it to a threshold: the model calculates the rate of change (or slope) of scores in key competency dimensions by comparing new and old feature vectors, and compares it with a pre-set, reasonable "improvement rate threshold" in the planning report. If the actual rate of change is consistently lower than the threshold, it indicates that the student is progressing slowly, may be encountering difficulties, or the plan may be out of touch with reality. At this point, the model will not wait for the end-of-term summary to detect the problem, but will trigger an early warning notification in real time. This warning is a proactive intervention signal that can prompt students to make self-adjustments or remind counselors to intervene and provide assistance in a timely manner, thereby resolving problems in their early stages and ensuring the effective implementation of career planning.

[0037] The above implementation method achieves a highly intelligent and adaptive student career planning approach. Its technical effectiveness is reflected in three aspects: First, the dynamic and forward-looking nature of the evaluation standards. By connecting with market data and adjusting model weights in real time, the evaluation results closely align with changes in social needs, overcoming the lag of traditional evaluations. Second, the accuracy and feasibility of the planning path. Based on quantified ability gaps, resource mapping and sequential planning are conducted, ensuring that the advice received by students is no longer vague but rather specific, orderly, and personalized action plans. Third, the closed-loop nature and timely intervention of the management process. Through continuous deviation monitoring and early warning, career planning is transformed from a static "one-off suggestion" into a dynamic "full-process growth support system," effectively improving the implementation effect of the plan and the sustainability of student development.

[0038] Reference Figure 2 As one implementation of step S102, the steps of cleaning and quantizing the original dataset using a standardization algorithm to generate feature vectors containing multi-dimensional capability indicators and outputting a standardized feature vector set include: Step S201: Perform data cleaning on the original dataset to remove duplicate, erroneous, or invalid data entries and generate a cleaned dataset. The raw dataset, derived directly from the preceding data collection steps, inevitably contains various data quality issues caused by system heterogeneity, manual input, or transmission errors. Data cleaning is an automated quality control process based on preset rules and algorithms.

[0039] Specifically, removing duplicate data entries aims to eliminate redundant records caused by multi-system synchronization or multiple entries, ensuring the unique representation of the same student behavior; removing erroneous data entries usually refers to identifying and correcting or eliminating obviously illogical values, such as course grades exceeding the full score or records with unreasonable date and timestamps; removing invalid data entries targets data with incorrect format, missing key fields (such as empty student IDs), or data marked as logically deleted.

[0040] It's important to note that this process is not simply a crude deletion, but rather involves contextual validation and imputation decisions. Technically, it utilizes data cleaning algorithms (such as rule-based validation and statistical outlier detection) to improve the integrity and consistency of the dataset, thereby producing a high-quality cleaned dataset.

[0041] Step S202: Based on the cleaned dataset, apply a quantization conversion algorithm to convert the unstructured capability data into numerical form, unify the data structure, and output a quantized dataset. The cleaned dataset still contains a large amount of "unstructured competency data," such as text descriptions in project reports, teacher comments, competition award levels, open-ended answers in psychological assessments, and performance evaluations in internship appraisals. The quantization transformation algorithm is the core technology for achieving this transformation. Its logic is to map non-numerical information into scalar values ​​based on pre-defined, domain-knowledge-driven scoring rules or models.

[0042] For example, for a text describing innovation, sentiment analysis or keyword extraction techniques from natural language processing can be used, combined with a predefined dictionary of innovative keywords and their weights, to calculate a score for "innovation ability." For ratings like "excellent," "good," and "qualified," a mapping rule (e.g., excellent = 90, good = 75) can be used to directly convert the data into numerical values. Simultaneously, a unified data structure means that the converted numerical values, along with the original structured numerical data (e.g., percentage grades), are reorganized according to a unified student-indicator-time-value record format, forming a well-organized quantitative dataset where all fields are numerical or can be directly used for numerical comparison.

[0043] Step S203: Extract preset multi-dimensional capability index data from the quantitative dataset, combine them to form a multi-dimensional feature vector, and generate a feature vector set; Specifically, based on the quantified dataset, the system performs extraction operations according to a predefined ability assessment framework (such as five dimensions: academic performance, innovation, social responsibility, mental health, and practical experience, with each dimension further subdivided). This is equivalent to constructing a feature vector for each student, which can be intuitively understood as a coordinate point of a student in the "ability space" spanned by these predefined ability dimensions.

[0044] Each dimension (i.e., each component) of the vector corresponds to a quantified value of a specific ability indicator. For example, the first component of the feature vector might represent "academic ability - GPA", the second component "innovation ability - project review score", the third component "social responsibility - volunteer service hours score", and so on. By selecting data from these specific dimensions in the quantified dataset and combining them in a fixed order, a student's unique ability feature vector is formed.

[0045] By performing this operation for all students, a feature vector set representing the entire group is generated. This feature vector set serves as the direct input for all subsequent mathematical models (such as calculating weighted scores, performing cluster analysis, and comparing differences).

[0046] Step S204: Standardize the feature vector set by adjusting it to a uniform numerical range using a scaling algorithm, and output the standardized feature vector set.

[0047] In the generated feature vector set, the values ​​for each dimension may come from vastly different sources: some from a percentage system of 0-100, some from a grading system of 1-5, and some from scores like 0.35 obtained through complex text analysis. Their numerical ranges and distribution variances vary greatly. If the original values ​​are used directly for calculation, indicators with larger numerical ranges will naturally dominate in subsequent weighted summation and other operations, thus distorting the true ability structure.

[0048] Therefore, normalization can be used in standardization. By employing a scaling algorithm (such as minimum-maximum normalization), a linear transformation is applied to all student values ​​in each dimension of the feature vector, mapping them as a whole to a preset uniform numerical range, most typically [0,1]. The principle is to subtract the minimum value among all values ​​in that dimension and then divide by the range of that dimension (maximum minus minimum). After this processing, the original score in each dimension is transformed into a relative position score between [0,1]. This eliminates the influence of the original units and only retains the student's relative level in this ability compared to all students.

[0049] Ultimately, in the output set of standardized feature vectors, the feature vector values ​​of each student are on the same scale across all dimensions. This allows for meaningful comparisons, additions, subtractions, and weighted calculations of scores across different dimensions, laying the foundation for the application of weights and the calculation of comprehensive scores in the subsequent dynamic evaluation model.

[0050] In the above implementation, the raw, heterogeneous, and multi-source student data collected from the educational information system is systematically transformed into a high-quality, structured, and standardized feature vector set with uniform scale. This provides a clean and directly mathematically operable standardized input for the upper-level dynamic multi-dimensional assessment model, fundamentally overcoming the shortcomings of traditional methods that suffer from insufficient reliability and validity of assessment results due to inconsistent data quality, varying formats, and mixed dimensions. Through this series of automated and algorithmic transformations, the objectivity, consistency, and computability of the student competency profile construction process are ensured, laying a reliable data foundation for subsequent accurate and dynamic career planning recommendations.

[0051] Reference Figure 3 As one implementation of step S103, the steps of obtaining market demand data through an external data interface, combining the preset professional competence matrix rules and market demand data, inputting the standardized feature vector set into a pre-constructed multi-dimensional competence assessment model for rule matching and weight coefficient update, calculating the competence assessment score, and outputting the competence assessment result set include: Step S301: Periodically acquire market demand data through an external data interface, parse the demand tags in the market demand data, and output a set of demand tags; The market demand data includes industry report data, enterprise job demand data, and graduate employment tracking data. Traditional evaluation models are often based on static and fixed indicator systems, which cannot respond to rapidly changing social and industry needs. This solution, however, establishes connections with external data sources through external data interfaces. These interfaces are programmable and can periodically capture or receive industry reports, recruitment website data, graduate employment quality reports, etc.

[0052] Specifically, the raw market demand data acquired is mostly unstructured text data, the core value of which lies in the descriptions of specific skills, qualities, and knowledge requirements. Therefore, keyword extraction, entity recognition, and semantic clustering techniques from natural language processing can be used to analyze the demand tags in the data. The system will identify frequently occurring and representative competency requirement terms (such as "Python programming," "cross-cultural communication," and "data analysis thinking") from massive amounts of text and standardize them into unified "demand tags."

[0053] Ultimately, the output demand tag set is a cleaned, deduplicated, and structured collection. It is essentially a machine-readable snapshot or profile of the current labor market's demand for talent capabilities, providing direct data input for subsequent adjustments to assessment criteria.

[0054] Step S302: Invoke the preset professional competence matrix rules, match the priority tags in the demand tag set with the core competence dimensions in the professional competence matrix rules, and generate a dynamic weight adjustment coefficient set; The generation of the dynamic weight adjustment coefficient set includes: allocating initial weights based on the matching degree between priority labels and core capability dimensions, and correcting historical weight coefficients through a time decay function.

[0055] Specifically, the pre-defined professional competency matrix rules are a structured knowledge framework defined by education experts. It clarifies which core competency dimensions (such as academic ability and innovation ability) the major aims to cultivate in students and their baseline importance. The matching operation is an algorithmic process based on rules or similarity calculations. The system performs semantic similarity calculations or rule mappings between each tag in the "demand tag set" and each core dimension in the competency matrix. For example, the market tag "machine learning application" may be highly correlated with the dimensions "innovation ability" and "practical ability."

[0056] More importantly, the system identifies priority tags, which are key needs determined by analyzing the frequency of tag occurrence, salary correlation, or industry trend. The matching result generates a set of dynamic weight adjustment coefficients.

[0057] Specifically, the algorithm calculates the change in the relative importance of each core competency dimension in the current market context based on the matching strength between labels and dimensions and the priority of the labels. A time decay function ensures that the adjustment is a smooth and gradual process; new market demands gradually influence the weights, while the influence of historical weights decays over time. This avoids drastic fluctuations in the evaluation criteria due to short-term market volatility, ensuring the stability and continuity of the evaluation. The final generated dynamic weight adjustment coefficient set is a set of weight parameters that will be applied to the evaluation of each student, reflecting the latest market expectations.

[0058] Step S303: Input the standardized feature vector set and the dynamic weight adjustment coefficient set into the pre-constructed multi-dimensional capability assessment model, and output the capability assessment score through the weighted calculation algorithm; The model input includes two key elements: one is the "standardized feature vector set" (each vector is a set of standardized scores for each dimension of the student) representing the student's current ability status, and the other is the "dynamic weight adjustment coefficient set" representing the current evaluation criteria preference.

[0059] In this embodiment, the main computational logic of the multi-dimensional ability assessment model is a "weighted calculation algorithm," the most common form of which is linear weighted summation, i.e., calculating the dot product of each student's feature vector and the dynamic weight vector. Specifically, the model multiplies the score representing a certain ability dimension (such as "practical ability") in the student's feature vector by the latest weight assigned to that dimension in the dynamic weight adjustment coefficient set, and then sums the weighted scores of all dimensions. This means that the same student's ability structure will yield different comprehensive ability assessment scores under different market weights.

[0060] For example, when the market highly values ​​"innovation ability," the weight of this dimension increases. Consequently, a student with strengths in innovation will see a corresponding rise in their overall score, more accurately reflecting their relative value in the current market environment. This calculation method transforms the assessment from static to a dynamic indicator reflecting the alignment between an individual's skill set and the demands of the times.

[0061] Step S304: Integrate all ability assessment scores to generate an ability assessment result set.

[0062] The competency assessment results set includes assessment scores for each dimension and a comprehensive score.

[0063] Specifically, following the parallel or iterative computation in the previous step, the system calculates one or more scores (potentially including scores for each dimension and a total score) for each individual student. A typical competency assessment result set is a data structure, such as a database table or a JSON array, with student identifiers as keys and records containing their competency assessment scores (total score and scores for each dimension) as values. This result set provides a quantitative evaluation of student competency based on the latest market standards.

[0064] In the above implementation, a dynamic adaptive system was constructed that can sensitively respond to changes in external market demand and automatically quantify these changes as adjustments to evaluation weights, which are then applied to the individual ability scores of each student. This technical solution overcomes the shortcomings of traditional student ability assessments, which are static, isolated, and disconnected from the professional world. It establishes an assessment standard that is linked to the labor market in real time and is continuously evolving. This ensures that the output ability assessment results are no longer merely academic measurements, but a dynamic assessment report of ability value rich in market context and with clear guiding significance, providing a solid decision-making foundation for subsequent highly targeted and forward-looking career planning.

[0065] Reference Figure 4 As one implementation of step S301, the steps of periodically acquiring market demand data through an external data interface, parsing the demand tags in the market demand data, and outputting a demand tag set include: Step S401: Periodically acquire market demand data through external data interfaces, including industry report data, enterprise job demand data, and graduate employment tracking data; The data in these reports typically comes from industry trend analyses published by authoritative institutions, used to capture the direction of skill evolution at the macro level; enterprise job demand data mainly comes from job descriptions on recruitment websites, reflecting in real time the specific skill requirements of enterprises for talent; and graduate employment tracking data comes from feedback from schools on the employment quality of previous graduates, used to verify the degree of matching between talent training and market demand. These three types of data together constitute a multi-source, heterogeneous market demand data pool, providing comprehensive data support for subsequent analysis and ensuring that the data sources of the evaluation system are broad and representative.

[0066] Step S402: Perform text preprocessing on the market demand data and output a standard text dataset; Specifically, most of the acquired market demand data exist in the form of unstructured natural language texts, which contain a large amount of redundant information and noise, and direct analysis will lead to a decrease in the accuracy of the results. Therefore, the system first performs invalid character removal processing on the data to eliminate HTML tags, special symbols and non-text content; then performs word segmentation processing, which uses a Chinese word segmentation algorithm to split the continuous text sequence into word units with independent semantics; and finally performs stop word filtering to remove high-frequency words that contribute little to semantics, such as "de", "le" and "zai".

[0067] Through the above preprocessing operations, the originally disordered raw text is converted into a standard text data set with clear structure and explicit semantics, which reduces the computational complexity of subsequent algorithms and improves data quality.

[0068] In step S403, a preset keyword extraction algorithm is called to analyze the standard text data set, extract candidate keywords related to industry competency requirements, calculate the term frequency and inverse document frequency of the candidate keywords in the standard text data set, and output a key competency word sequence; This step is implemented using natural language processing technology by calculating the term frequency (TF) and inverse document frequency (IDF) of the candidate keywords in the standard text data set. Term frequency reflects the occurrence frequency of a certain word in a single document, and a higher occurrence frequency usually means higher importance; while inverse document frequency measures the prevalence of the word in the entire data set. If a certain word appears in multiple documents, its discrimination degree decreases, and the IDF value decreases accordingly.

[0069] By combining the two (i.e., the TF-IDF algorithm), the system can effectively identify keywords that appear frequently in specific documents but are relatively sparse in the entire data set. These keywords often represent the industry's current urgently needed and distinctively featured competency requirements, so as to output a highly representative key competency word sequence.

[0070] In step S404, a preset competency standard vocabulary database is acquired, the semantic similarity between the key competency word sequence and each standard vocabulary in the competency standard vocabulary database is calculated, key competency words with a similarity higher than a preset threshold are mapped to standard competency tags, and an original tag set is output; Specifically, in actual texts, the same ability may be expressed in multiple ways. For example, "programming ability" and "code writing ability" are semantically highly consistent but literally different. To achieve accurate matching, the system pre-defines a unified standard vocabulary for abilities and uses semantic similarity calculation techniques (such as cosine similarity calculation based on word vectors) to calculate the semantic similarity between the extracted key ability word sequences and each standard word in the standard vocabulary. When the semantic similarity between the two is higher than a preset threshold, the system determines that they refer to the same ability concept, thus mapping the non-standardized key ability words to standardized ability labels and outputting the original label set. This step effectively eliminates data redundancy caused by synonyms and near-synonyms, ensuring the uniformity and comparability of the evaluation dimensions.

[0071] Step S405: Count the frequency of occurrence of each standard capability tag in the original tag set, sort the standard capability tags in descending order of frequency, select the first preset number of tags as requirement tags, and integrate them to obtain the requirement tag set.

[0072] The system performs frequency statistics on all mapped standard capability tags. The frequency of each tag's occurrence directly reflects the current market demand for that capability. Sorting tags by frequency from highest to lowest provides a clear picture of market demand for various capabilities. The system selects a predetermined number of tags from the top of the sorted list as the final demand tags and integrates them to obtain a demand tag set. This process extracts core structured indicators from massive amounts of unstructured data, ensuring that the evaluation dimensions used in subsequent capability assessment models closely align with cutting-edge market demands and capture the core characteristics of dynamic industry changes.

[0073] The above implementation achieves in-depth mining and intelligent processing of market demand data, automatically extracting standardized competency requirement tags from massive amounts of heterogeneous unstructured text and prioritizing them according to demand popularity. This not only solves the problems of lag and subjectivity caused by the reliance on manually set indicators in traditional assessment systems, but also ensures that the dynamic update mechanism of competency assessment dimensions has solid data support, enabling students' competency assessment results to respond in real time to rapid changes in the labor market, and improving the scientific nature and timeliness of career planning.

[0074] Reference Figure 5 As one implementation of step S104, the steps of generating a capability distribution radar chart and automatically generating an analysis report based on the capability assessment result set, and outputting the visualization charts and analysis report, include: Step S501: Based on the capability assessment result set, extract capability score data for each dimension and generate a radar chart data input set; The input ability assessment result set is a structured data object (such as JSON or a data table) containing student identification, ability scores in various dimensions (such as academic 85, innovation 70, practice 90, etc.), and possible comprehensive scores.

[0075] However, general-purpose visualization engines (such as ECharts and D3.js) need to adhere to their specific data format protocols when rendering radar charts. Extracting ability score data for each dimension is a selective reading and reorganization process performed in a predetermined dimensional order (e.g., [academics, innovation, social responsibility, mental health, practical experience]). This is not merely about simply retrieving numbers, but more importantly, about constructing a data structure that conforms to the radar chart coordinate system.

[0076] In this embodiment, the generated radar chart data input set is a paired object containing a list of dimension names and a list of corresponding score values. For example, it explicitly defines the positional order of each axis (dimension) and its corresponding length (score) on the two-dimensional plane of the radar chart.

[0077] Step S502: Call the visualization engine to render the radar chart data input set into a polygon radar chart and set visual distinguishing labels to represent different capability states. The visual differentiation markers include color coding or line differentiation, which are used to identify the current ability level and the target ability level.

[0078] Specifically, the visualization engine is a software library that performs graphic rendering. It receives the data input set prepared in the previous step. Its internal algorithm, based on the mathematical principles of radar charts, maps the score of each dimension to a point on an axis emanating from the center of the circle in a polar coordinate system. Then, it connects the points in all dimensions sequentially to form a closed polygon. The geometric shape of this polygon is a visual portrait of the student's ability structure: the larger the area and the more outward the outline, the stronger or more balanced the comprehensive ability; the prominence or indentation of a vertex clearly indicates strengths and weaknesses.

[0079] Next, visual markers are set up to represent different ability states. Richer comparative information can be conveyed through the overlay of visual codes. The most common approach is to use "solid-line polygons" to represent a student's current actual level, and "dashed-line polygons" or overlays of polygons of different colors to represent the target level or industry benchmark level. Color coding (e.g., using green to indicate meeting the standard and red to indicate not meeting it) or different line types transform the abstract concept of "gap" into visual signals that can be understood without conscious thought, greatly reducing the cognitive load of information interpretation.

[0080] Step S503: Based on the competency assessment result set, automatically parse the competency score data using a report generation algorithm and output the analysis report content; The core logic of this step is automatic reasoning and text generation from numerical values ​​to semantics, which aims to supplement the intuitive perception provided by radar charts with accurate qualitative descriptions and action-oriented textual interpretations.

[0081] Specifically, the report generation algorithm first parses the ability score data, which typically involves a series of rule-based logical judgments or simple pattern recognition: for example, sorting the scores of each dimension to identify the highest score (strength dimension) and the lowest score (weakness dimension); comparing the score of each dimension with a preset threshold or target value to determine whether it is in the "excellent", "acceptable" or "needs improvement" range; and calculating the score difference between different dimensions to assess the balance of the ability structure.

[0082] Next, the algorithm selects descriptive fragments, evaluation statements, and suggestions that match the above judgment results from a pre-set text template library rich in domain knowledge, and performs intelligent splicing and variable filling. For example, if it identifies that the "innovation ability" score is outstanding and higher than the target value, it selects statements such as "demonstrates outstanding innovative thinking and problem-solving ability" from the "strengths description library"; if it identifies that the "practical ability" score is significantly low, it calls specific suggestions such as "suggests participating in at least one professionally related internship or project practice" from the "improvement suggestion library".

[0083] Ultimately, the output analysis report is a coherent, personalized, and directly readable natural language text, which completes the transformation from cold numbers to warm, actionable guidance.

[0084] Step S504: Integrate the polygon radar chart and analysis report content to output visualization charts and analysis reports.

[0085] After the first two independent processing steps, the system has generated two different outputs describing the same object: a graphical "polygon radar chart" and a textual "analysis report content." Integration means organically linking and unifying these two at the system or presentation level by embedding the graph and text in specific locations within the same webpage view or PDF document, ensuring they both refer to the same student. Furthermore, deeper integration includes interactive linkages; for example, when the report mentions "As shown in the radar chart, your innovation ability is particularly outstanding," the corresponding "innovation" dimension on the radar chart can be highlighted.

[0086] Ultimately, the output visualizations and analysis reports constitute a complete and integrated analytical product. The charts provide a global overview, intuitive comparisons, and quick location of information, while the report offers detailed descriptions, causal analysis, and specific recommendations. The two complement and corroborate each other, forming a complete cognitive loop from macro to micro, from phenomena to suggestions, enabling users to efficiently and comprehensively understand their evaluation results, whether through rapid visual scanning or in-depth text reading.

[0087] The above implementation constructs a system that automatically transforms multidimensional assessment data into comprehensive insight reports that combine visual intuitiveness with semantic depth. Through multidimensional comparison of radar charts and visual identifiers, rapid insight into assessment conclusions is achieved. Simultaneously, rule-based automated text generation provides personalized and actionable in-depth interpretations. This not only enhances the comprehensibility and user acceptance of the results but also replaces manually written reports with standardized algorithm output, ensuring the feasibility of scalable analysis and the consistency of conclusions. This allows the results of dynamic assessments to reach every student in the most user-friendly and effective way, providing strong dual support for their self-awareness and planning decisions.

[0088] Reference Figure 6 As one implementation of step S105, the steps of receiving a development direction selection instruction input from the user interface, applying a path planning algorithm to compare ability differences and map resources, generating personalized career planning suggestions and progressive ability improvement paths, and outputting a personalized planning report include: Step S601: Receive the development direction selection instruction input from the user interface, parse the development direction type in the development direction selection instruction and output the development direction identifier; wherein, the development direction type includes further education or employment; In this embodiment, the development direction selection made by the user through the graphical user interface (such as clicking the "Further Education" or "Employment" button) is a decision expression based on a preliminary synthesis of their personal aspirations, interests, and external information. After receiving this input, the system performs semantic analysis to extract and standardize key intentions.

[0089] For example, if a user selects "Employment - Internet Technology", the parser will convert it into a machine-readable, internally defined career direction identifier. This identifier anchors the user's free and diverse natural language or interactive choices to a pre-set, algorithm-processable structured target framework.

[0090] Step S602: Query the preset target capability database according to the development direction identifier to obtain the target capability score, compare the capability assessment score in the capability assessment result set with the target capability score, calculate the capability gap value and generate a capability difference dataset. The system retrieves a detailed target competency database from its internal knowledge base based on the development direction identifier. This database contains quantitative standards (target scores) for each dimension of competency required to succeed in that development direction (such as being competent for the target position or meeting postgraduate admission requirements). Simultaneously, the system accesses a competency assessment result set representing the student's current status, which includes the student's actual competency scores for each dimension.

[0091] Next, the ability assessment scores representing the current status of students are compared with the target ability scores representing development goals through element-by-element subtraction or similarity calculation (such as Euclidean distance or cosine similarity). The comparison result is a structured ability difference dataset, which clearly lists the student's current score, target score, and the gap between the two in each ability dimension. It indicates which dimensions are strengths (negative or zero gap) and which are weaknesses that urgently need to be improved (positive and large gap), providing precise input and priority ranking basis for the next step of resource matching.

[0092] Step S603: Input the capability difference dataset into the preset resource mapping rule base for matching, obtain recommended resources corresponding to the capability gap value, and output the resource matching result set; wherein, the resource mapping rule base includes training course data, practical activity data and career development path templates; In this embodiment, the resource mapping rule base is a knowledge base or rule engine that stores the "capability-resource" relationship. The recommended resources include all intervention measures that the system can schedule, such as specific training courses, online micro-professions, internship projects, competitions, certification exams, books and materials, etc. Each resource is labeled with one or more specific capability dimensions that it can effectively improve and the estimated improvement effect.

[0093] Specifically, based on each positive gap (i.e., weakness) identified in the ability difference dataset, the rule base is queried and filtered to find one or more resources that can most effectively fill that gap. For example, for a significant gap in the "Data Structures and Algorithms" ability dimension, the rule base might map to an online course on "Introduction to Algorithms," a LeetCode problem-solving program, and a recommended algorithm competition. This process is not a simple one-to-one matching; it may involve complex many-to-many relationships and requires intelligent sorting and combination based on factors such as the difficulty level of resources, their priority, and time costs. The final resource matching result set is a feasible list of resources customized for the student's specific ability weakness.

[0094] Step S604: Based on the resource matching result set, arrange the recommended resources according to the preset ability improvement stage order to generate a progressive ability improvement path sequence, and generate personalized career planning suggestions according to the development direction identifier. In the embodiments of this application, the path planning algorithm needs to comprehensively consider a variety of constraints, such as the prerequisites for knowledge points (programming basics must be learned before machine learning can be learned), the time availability of resources (summer internships, fall semester courses), the balance of learning load, and the phased goals of personal development.

[0095] Based on these constraints, the algorithm will perform resource scheduling and sorting to generate a progressive sequence of capability improvement paths. This sequence will be organized into pre-defined capability improvement stages such as "short-term, medium-term, and long-term." For example, the short-term (this semester) focus on supplementing core theoretical knowledge through two online courses; the medium-term (summer vacation) involves arranging a relevant internship project to accumulate practical experience; and the long-term (next academic year) plan includes obtaining an advanced certification and participating in competitions.

[0096] At the same time, the algorithm will generate corresponding personalized career planning suggestion texts for each stage and key resource node, explaining why this arrangement is made and the expected goals.

[0097] Step S605: Integrate the progressive skills enhancement path sequence with personalized career planning suggestions, and output a personalized planning report.

[0098] Specifically, the system has generated two main parts: first, descriptive and explanatory personalized career planning suggestions; and second, a structured, timeline-based or checklist-style progressive skills enhancement path sequence. Integration means organically merging and arranging the content and presentation, for example, embedding the path sequence into the report in the form of Gantt charts, timelines, or milestone lists, and explaining, emphasizing, and encouraging it with planning suggestions in appropriate places.

[0099] Ultimately, the resulting personalized planning report is usually presented in the form of a document or an interactive page. It begins by clearly stating the development goals, then analyzes the capability gaps, then proposes overall recommendations, and finally lists in detail the specific action steps (path sequence) in stages, which are traceable and time-bound, with relevant resource links or instructions attached next to each step.

[0100] The above implementation provides an intelligent planning system that automatically generates concrete and actionable action plans based on the user's subjective intentions. Driven by precise capability difference analysis, it not only identifies the ultimate goal but also draws a detailed roadmap and provides necessary recommended resources. This transforms career development planning from static consultation advice into a dynamic, data-driven, and traceable intelligent navigation service, enhancing the scientific rigor, personalization, and operability of the planning.

[0101] Reference Figure 7As a further implementation of the dynamic multi-dimensional student ability assessment and career planning method, after step S105, which generates personalized career planning suggestions and progressive ability improvement paths and outputs a personalized planning report, the method further includes: Step S701: Analyze the personalized planning report, extract the progressive capability improvement path sequence, and determine the recommended resource set corresponding to each capability improvement stage; The progressive capability enhancement path sequence is a list of action plans organized by time phases (such as short-term, medium-term, and long-term). Each plan item not only contains capability enhancement goals, but also associates specific resources recommended to achieve those goals (such as a course, an internship project, or a certification exam) through resource mapping relationships.

[0102] In this embodiment, the system uses a parsing algorithm to traverse each stage and each target node of the path sequence, and, based on its built-in resource mapping rules, makes explicit and list the implicit or explicit resource dependencies. For example, for the stage goal of "improving data analysis capabilities," the mapping relationship may point to "taking the 'Statistical Machine Learning' course" and "completing a Kaggle competition project." After parsing, a set of requirements containing these two specific resources is generated.

[0103] Ultimately, the output recommended resource set is a structured, phased resource list that accurately depicts the "bill of materials" of the original planning scheme at the resource level, providing a clear benchmark and optimization targets for subsequent evaluation of the availability and applicability of these resources in the current environment.

[0104] Step S702: Obtain dynamic market demand change data and education platform resource update data in real time through external data interface to generate dynamic resource status dataset; The system obtains data from two key sources through pre-defined external data interfaces, using either periodic polling or event listening: first, "dynamic market demand data," which typically comes from recruitment website APIs, industry research report databases, macroeconomic indicators, etc., to perceive the rise and fall of skill demands, the emergence of new positions, and the evolution of talent competency requirements; second, "education platform resource update data," which comes from school academic affairs systems, MOOC platforms, internship information databases, competition activity publishing platforms, etc., to keep track of real-time status such as course start and stop, internship postings, training program updates, and activity registration deadlines.

[0105] Subsequently, the collected raw streaming data was cleaned, deduplicated, and formatted before being integrated to generate a dynamic resource status dataset. This dataset not only records the static attributes of resources (such as name and description), but more importantly, it includes their dynamic status (such as availability, quota remaining, deadline, and latest version) and related tags associated with market demand trends.

[0106] Step S703: Invoke the preset resource matching algorithm to perform matching calculations on the recommended resource set and the dynamic resource status dataset, and generate a resource optimization mapping table containing resource availability scores and alternative resource identifiers; The core logic of this step lies in using an algorithmic model to find the optimal and feasible matching solution between ideal planning requirements and realistic resource constraints.

[0107] Specifically, the resource matching algorithm is the core technology for performing this complex multi-objective optimization. Its input consists of a set of recommended resources representing "ideal demand" (first step output) and a dynamic dataset of resource status representing "actual supply" (second step output). The algorithm's working mechanism involves multiple calculations and evaluations: first, availability verification is performed, checking whether each resource in the recommended resource set exists in the dynamic dataset and whether it is in an accessible state (e.g., courses are available during the enrollment period, internship positions are open for recruitment).

[0108] Secondly, a fit score is performed. Even if a resource is available, its quality of matching with the original requirement goal needs to be evaluated. For example, a newly offered "Python Data Analysis" course might have a 90% match with the requirement goal of "mastering data processing skills." This score is called the "resource availability score," a quantitative indicator that integrates multiple factors such as timeliness, cost, and prerequisites. More importantly, when the original required resource is unavailable or has a low score, the algorithm initiates an alternative resource discovery process. Based on rules such as semantic similarity and equivalence of capability enhancement goals, it searches for functionally similar or better alternatives in the dynamic dataset and labels them with "alternative resource tags."

[0109] Ultimately, the generated resource optimization mapping table clearly lists the evaluation results of each original resource: whether it is "healthy and available" (high score), "requires enabling alternatives" (with alternative identifier), or "no suitable resource available" (low score and no alternative).

[0110] Step S704: Based on the resource optimization mapping table, perform an update operation on the recommended resource allocation scheme in the progressive capability enhancement path sequence, and output the optimized capability enhancement path sequence; In this embodiment, the system uses a resource optimization mapping table as an operation guide to dynamically update the original "progressive capability enhancement path sequence". This is essentially a rule-based editing process: resources marked with high availability scores in the mapping table are retained and confirmed in the path sequence; for items marked with alternative resource identifiers, the system automatically replaces the resource node in the original path with the recommended alternative resource; for resources with extremely low scores and no alternatives, the target node may be marked as "pending" or a higher-level warning may be triggered.

[0111] Understandably, the entire update process follows the principle of "unchanged goals, optimized paths," meaning that the capability enhancement goals at each stage remain unchanged, but the "tools" or "paths" (i.e., resources) for achieving those goals are optimized based on the actual situation. For example, the original plan of "participating in an offline internship at Company A during the summer" was marked as unavailable because all spots were filled, while the mapping table recommended "a remote internship program at Company B" as an alternative. After the update, this node in the path sequence is replaced. The output "optimized capability enhancement path sequence" is a revised action plan that is synchronized in real time with the current resource environment and has enhanced feasibility.

[0112] Step S705: Obtain the planning execution progress data through the student behavior monitoring interface, combine it with the optimized capability improvement path sequence, and call the pre-built execution deviation analysis model to calculate the path execution deviation coefficient; The logic behind this step is to continuously compare students' actual behavior with the optimized plan baseline, and use mathematical models to accurately measure the degree of deviation in the execution process, thereby achieving a leap from qualitative observation to quantitative diagnosis.

[0113] Specifically, the system continuously acquires planning and execution progress data from sources such as the learning management system, project collaboration platform, and assessment system through student behavior monitoring interfaces. This data includes course completion status, project milestone completion, skills certification acquisition, and time invested in learning. Simultaneously, the core task of the execution deviation analysis model is to calculate the "path execution deviation coefficient" by analyzing multiple dimensions: time deviation (the offset between the actual completion date and the planned date), schedule deviation (the difference between the current completed workload and the expected percentage of total planned workload), and effect deviation (the gap between the actual improvement in competency value brought about by completed tasks and the expected improvement). By weighted fusion of these dimensions, the model calculates a normalized, comparable deviation coefficient. This coefficient is a continuous value, and its magnitude directly reflects the overall degree to which the actual execution trajectory deviates from the ideal planned trajectory, providing a precise, data-driven basis for determining whether intervention is necessary.

[0114] Step S706: When the path execution deviation coefficient is detected to exceed the preset execution deviation threshold, a path adjustment trigger command is generated; The preset execution deviation threshold is a configurable parameter that defines the upper limit of the planned execution deviation that the system can tolerate. This threshold can be set differently according to the importance of different capability dimensions and the rigor of different planning stages.

[0115] In this embodiment, deviation detection is a continuous judgment process. When the coefficient calculated by the deviation analysis model continuously or instantaneously exceeds the threshold, the system determines that a "significant deviation" has occurred and system intervention is required. At this time, the system generates a path adjustment trigger command. This control signal encapsulates the core information of the deviation event (such as student ID, stage of deviation, deviation coefficient value, and main deviation dimension).

[0116] Step S707: In response to the path adjustment trigger command, the path reconstruction algorithm is invoked to regenerate the progressive capability enhancement path sequence based on the resource optimization mapping table and the capability assessment result set, and an updated personalized planning report is output.

[0117] Upon receiving the path adjustment trigger command, the system invokes a more powerful path reconstruction algorithm. This algorithm is no longer limited to local repairs of the original path, but instead uses the "current resource optimization mapping table" (reflecting the latest resource environment) and the "ability assessment result set" (reflecting the student's latest ability status at this moment) as core inputs to recalculate the path planning. Starting from the student's current state and ending with the original or adjusted ultimate development direction, it recalculates an optimal progressive ability improvement path sequence under the latest resource constraints.

[0118] Understandably, this process comprehensively considers all available resources, new time constraints, and problems exposed in previous execution deviations (such as low efficiency in certain types of tasks) to generate a new plan that is more feasible and better suited to the student's current situation. The final output, an updated personalized planning report, serves as a new guide for the student's next stage of action, thus initiating a new "execution-monitoring-optimization" cycle.

[0119] In the above implementation, the system can perceive changes in external resources and market demand in real time and continuously monitor the internal execution status. It can proactively prevent planning interruptions caused by resource failures at the front end and promptly correct path deviations caused by execution deviations at the back end. The system can replan the path based on the latest full information, ensuring that the career development path provided to students is always the most feasible, optimized, and suitable choice for their personal progress status in the current environment, thereby improving the achievement rate of planning goals and the sustainability of student development.

[0120] Reference Figure 8As one implementation of step S106, the steps of acquiring newly collected student ability data based on the personalized planning report, calling the deviation detection model to monitor the deviation in student ability improvement progress, and triggering an early warning notification when the deviation exceeds a threshold include: Step S801: Obtain the progressive capability improvement path sequence from the personalized planning report and extract the set of stage target capability indicators from the planning path; The progressive capability enhancement path sequence not only includes the actions to be performed (such as attending a course), but also implicitly includes the capability enhancement results expected to be achieved by each action, namely the "stage target capability indicators". The system automatically parses the path sequence and identifies all explicit or implicit capability target nodes.

[0121] For example, for a plan item in the path sequence such as "Complete the course of 'Machine Learning Fundamentals' at the end of the first semester, with the goal of improving the ability value of the algorithm understanding dimension to 75 points", after extraction, a "target ability index" (target value 75 points) will be generated for the algorithm understanding dimension at the time point of "end of semester".

[0122] Ultimately, the generated set of stage-specific target ability indicators is a structured dataset that clearly defines the expected levels of students' various abilities to be achieved on the planned timeline, according to time stages (such as short-term, medium-term, and long-term) and ability dimensions.

[0123] Step S802: Continuously acquire newly collected student ability data through the data acquisition interface to generate a real-time ability dataset; Specifically, the system automatically acquires newly collected student ability data from academic affairs, practical training, and assessment systems through existing data acquisition interfaces, according to preset cycles (such as monthly or per semester) or event triggers (such as when a student completes a course or a practical activity). After necessary cleaning and standardization, this data is organized into a real-time ability dataset with a structure consistent with the "stage target ability indicator set". This dataset reflects the actual latest values ​​of students' abilities in each dimension at the current monitoring point in time.

[0124] Step S803: Call the deviation detection model, compare the real-time capability dataset with the stage target capability index set, and calculate the progress deviation value of each capability dimension. The deviation detection model is input to a real-time capability dataset representing the "actual trajectory" and a set of stage-specific target capability indicators representing the "planned baseline." This model focuses not only on the absolute difference between the current capability value and the stage-specific target value, but also on whether the rate or trend of capability improvement deviates from expectations.

[0125] For example, the model may calculate the actual growth rate of a certain capability value from the previous monitoring point to the current point and compare it with the average growth rate required to achieve the stage target as preset in the plan. The sliding window algorithm can be used to analyze multiple data points in the most recent time window to fit the capability change trend line, thereby smoothing the fluctuation of single-point data and more stably judging whether the trend deviates.

[0126] Ultimately, the model outputs a quantified progress deviation value for each monitored capability dimension. This value can be a percentage (such as the actual improvement rate being only 60% of the expected rate) or a standardized score.

[0127] Step S804: When the progress deviation value of at least one capability dimension exceeds the preset deviation threshold, an early warning event identifier is generated. The calculated schedule deviation is a continuous variable, and the system needs a clear rule to determine whether it constitutes a problem requiring intervention. The preset deviation threshold is the key parameter of this decision rule. The preset deviation threshold can be set differently according to the importance (dynamic weight) of different capability dimensions. The threshold for important capabilities (such as core professional skills) may be stricter (e.g., a warning is issued if the deviation exceeds 5%), while the threshold for auxiliary capabilities may be more lenient (e.g., a warning is issued only if the deviation exceeds 15%).

[0128] Specifically, when the deviation value of any dimension output by the deviation detection model exceeds the set threshold, it means that there has been an abnormal deviation in that dimension. At this time, the system will generate an early warning event identifier, which includes at least the following key information: which student, which ability dimension, at what time, what type (e.g., progress delay) and what degree (deviation value) of deviation occurred.

[0129] Step S805: Based on the warning event identifier, trigger the warning notification push service to the user terminal.

[0130] The final generated warning event identifier is an internal system signal that needs to be delivered to the student or their advisor. Triggering the warning notification push service means that the system calls an external service or module specifically responsible for message distribution. This service receives the warning event identifier, and based on its content (such as the severity of the deviation and the capabilities involved) and predetermined rules (such as tiered push rules), generates specific and readable "warning notification" content, and determines its push channel (such as system in-app message, email, SMS) and recipients (such as students, counselors, and tutors).

[0131] For example, a minor delay might only trigger a reminder for the student; a serious or multi-dimensional persistent deviation might be simultaneously pushed to both the student and the teacher. In some embodiments, the final output alert notification is a specific message, such as "Tip: Your 'project practice ability' improvement is less than 20% of the planned progress. Please pay attention and adjust your study plan."

[0132] In the above implementation, long-term planning goals are broken down into monitorable stage indicators, and continuous data streams are used for real-time comparison and trend analysis. This allows for the automatic and accurate identification of risks and the issuance of early warnings when deviations occur. This data- and algorithm-based dynamic monitoring mechanism not only improves the compliance and success rate of planning implementation but also transforms career counseling from a passive, remedial approach to proactive, preventative support, achieving refined and intelligent accompanying support for students' growth process.

[0133] This application also discloses a dynamic multi-dimensional student ability assessment and career planning system.

[0134] A dynamic, multi-dimensional student ability assessment and career planning system, specifically including: The multi-source data integration and consolidation module is used to collect structured and unstructured student ability data from the education information system, integrate the data, and output the original dataset. The standardized vector generation module is used to clean and quantize the original dataset by calling a standardization algorithm, generate feature vectors containing multi-dimensional capability indicators, and output a standardized feature vector set. The dynamic capability assessment module is used to obtain market demand data through external data interfaces, combine the preset professional capability matrix rules and market demand data, input the standardized feature vector set into the pre-built multi-dimensional capability assessment model for rule matching and weight coefficient update, calculate the capability assessment score and output the capability assessment result set; The visualization analysis module is used to generate a capability distribution radar chart based on the capability assessment result set, and automatically generate an analysis report, outputting visualization charts and analysis reports. The personalized path planning module receives development direction selection instructions from the user interface, applies path planning algorithms to compare ability differences and map resources, generates personalized career planning suggestions and progressive ability improvement paths, and outputs a personalized planning report. The progress monitoring and feedback module is used to acquire newly collected student ability data based on personalized planning reports, call the deviation detection model to monitor deviations in student ability improvement progress, and trigger an early warning notification when the deviation exceeds a threshold.

[0135] The dynamic multi-dimensional student ability assessment and career planning system of this application embodiment can implement any of the above methods, and the specific working process of each module in the system can refer to the corresponding process in the above method embodiment.

[0136] In the several embodiments provided in this application, it should be understood that the provided methods and systems can be implemented in other ways. For example, the system embodiments described above are merely illustrative; for example, the division of a certain module is merely a logical functional division, and in actual implementation there may be other division methods, such as multiple modules can be combined or integrated into another system, or some features can be ignored or not executed.

[0137] This application also discloses a computer-readable storage medium.

[0138] A computer-readable storage medium storing a computer program that can be loaded by a processor and executed as described above in any of the methods of a dynamic multidimensional student ability assessment and career planning approach.

[0139] The computer-readable storage medium can be any tangible medium that contains or stores a program that can be used by or in connection with an instruction execution system, apparatus, or device; the program code contained on the computer-readable medium can be transmitted using any suitable medium, including but not limited to wireless, wire, optical fiber, RF, etc., or any suitable combination thereof.

[0140] The above are all preferred embodiments of this application and are not intended to limit the scope of protection of this application. Any feature disclosed in this specification (including the abstract and drawings) may be replaced by other equivalent or similar features unless specifically stated otherwise. That is, unless specifically stated otherwise, each feature is only one example of a series of equivalent or similar features.

Claims

1. A dynamic, multi-dimensional student ability assessment and career planning method, characterized in that, The method includes: Structured and unstructured student ability data are collected from the education information system, and the raw dataset is output after data integration. The original dataset is cleaned and quantized using a standardization algorithm to generate feature vectors containing multi-dimensional capability indicators, and a standardized feature vector set is output. Market demand data is obtained through an external data interface. Combined with the preset professional competence matrix rules and the market demand data, the standardized feature vector set is input into a pre-constructed multi-dimensional competence assessment model for rule matching and weight coefficient update. The competence assessment score is calculated and the competence assessment result set is output. Based on the capability assessment results set, the visualization engine is invoked to generate a capability distribution radar chart and automatically generate an analysis report, outputting visualization charts and analysis reports; It receives development direction selection instructions from the user interface, applies path planning algorithms to compare ability differences and map resources, generates personalized career planning suggestions and progressive ability improvement paths, and outputs a personalized planning report. Based on the personalized planning report, newly collected student ability data is obtained, and a deviation detection model is called to monitor the deviation in the progress of student ability improvement. When the deviation exceeds the threshold, an early warning notification is triggered.

2. The method for dynamic multi-dimensional student ability assessment and career planning according to claim 1, characterized in that, The steps involved in cleaning and quantizing the original dataset using a standardization algorithm to generate feature vectors containing multidimensional capability indicators and outputting a standardized feature vector set include: Perform data cleaning operations on the original dataset to remove duplicate, erroneous, or invalid data entries and generate a cleaned dataset. Based on the cleaned dataset, a quantization conversion algorithm is applied to convert the unstructured capability data into numerical form, unify the data structure, and output a quantized dataset. Extract preset multi-dimensional capability index data from the quantified dataset, combine them to form a multi-dimensional feature vector, and generate a feature vector set. The feature vector set is standardized by scaling it to a uniform numerical range, and a standardized feature vector set is output.

3. The method for dynamic multi-dimensional student ability assessment and career planning according to claim 1, characterized in that, The steps include: acquiring market demand data through an external data interface, combining the market demand data with preset professional competence matrix rules, inputting the standardized feature vector set into a pre-constructed multi-dimensional competence assessment model for rule matching and weight coefficient updates, calculating competence assessment scores, and outputting a competence assessment result set. Market demand data is periodically acquired through external data interfaces, the demand tags in the market demand data are parsed, and a set of demand tags is output. Invoke the preset professional competence matrix rules, match the priority tags in the demand tag set with the core competence dimensions in the professional competence matrix rules, and generate a dynamic weight adjustment coefficient set; The standardized feature vector set and the dynamic weight adjustment coefficient set are input into a pre-constructed multi-dimensional capability assessment model, and the capability assessment score is output through a weighted calculation algorithm. Integrate all ability assessment scores to generate an ability assessment result set.

4. The dynamic multi-dimensional student ability assessment and career planning method according to claim 3, characterized in that, The steps of periodically acquiring market demand data through an external data interface, parsing the demand tags in the market demand data, and outputting a set of demand tags include: We periodically acquire market demand data through external data interfaces, including industry report data, enterprise job demand data, and graduate employment tracking data. Perform text preprocessing on the market demand data to output a standard text dataset; The pre-defined keyword extraction algorithm is invoked to analyze the standard text dataset, extract candidate keywords related to industry capability requirements, and calculate the word frequency and inverse document frequency of the candidate keywords in the standard text dataset, outputting a key capability word sequence. Obtain a preset capability standard vocabulary library, calculate the semantic similarity between the key capability word sequence and each standard word in the capability standard vocabulary library, map key capability words with similarity higher than a preset threshold to standard capability labels, and output the original label set; The frequency of occurrence of each standard capability tag in the original tag set is counted. The standard capability tags are sorted in descending order of frequency. A preset number of tags at the top of the sorted list are selected as requirement tags, and the requirement tag set is obtained by integrating them.

5. The dynamic multi-dimensional student ability assessment and career planning method according to claim 1, characterized in that, Based on the capability assessment results set, the steps of generating a capability distribution radar chart and automatically producing an analysis report by calling a visualization engine include: Based on the competency assessment result set, extract the competency score data for each dimension and generate a radar chart data input set; The visualization engine is invoked to render the radar chart data input set into a polygon radar chart, and visual distinguishing labels are set to represent different capability states. Based on the competency assessment results set, the competency score data is automatically parsed using a report generation algorithm, and the analysis report content is output. Integrate polygon radar charts and analysis report content to output visual charts and analysis reports.

6. The dynamic multi-dimensional student ability assessment and career planning method according to claim 1, characterized in that, The steps involved in receiving development direction selection instructions from the user interface, applying path planning algorithms to compare ability differences and map resources, generating personalized career planning suggestions and progressive ability improvement paths, and outputting a personalized planning report include: The system receives a development direction selection instruction from the user interface, parses the development direction type in the instruction, and outputs a development direction identifier; wherein, the development direction type includes further education or employment. The target capability score is obtained by querying the preset target capability database according to the development direction identifier. The capability assessment score in the capability assessment result set is compared with the target capability score to calculate the capability gap value and generate a capability difference dataset. The capability difference dataset is input into a pre-set resource mapping rule base for matching to obtain recommended resources corresponding to the capability gap value, and a resource matching result set is output; wherein, the resource mapping rule base includes training course data, practical activity data and career development path templates; Based on the resource matching result set, the recommended resources are arranged in a preset order of capability improvement stages to generate a progressive capability improvement path sequence, and personalized career planning suggestions are generated based on the development direction identifier. Integrate the progressive skills enhancement path sequence with the personalized career planning suggestions to output a personalized planning report.

7. The dynamic multi-dimensional student ability assessment and career planning method according to claim 6, characterized in that, After generating personalized career planning suggestions and progressive skills enhancement paths, and outputting a personalized planning report, the process also includes: The personalized planning report is analyzed to extract the progressive capability improvement path sequence and determine the recommended resource set corresponding to each capability improvement stage. By acquiring real-time data on dynamic changes in market demand and updates to educational platform resources through external data interfaces, dynamic resource status datasets are generated. A preset resource matching algorithm is invoked to perform matching calculations on the recommended resource set and the dynamic resource status dataset, generating a resource optimization mapping table that includes resource availability scores and alternative resource identifiers; Based on the resource optimization mapping table, the recommended resource allocation scheme in the progressive capability enhancement path sequence is updated, and the optimized capability enhancement path sequence is output. The plan execution progress data is obtained through the student behavior monitoring interface, and the path execution deviation coefficient is calculated by calling the pre-built execution deviation analysis model in combination with the optimized capability improvement path sequence. When the path execution deviation coefficient is detected to exceed the preset execution deviation threshold, a path adjustment trigger command is generated; In response to the path adjustment trigger command, the path reconstruction algorithm is invoked to regenerate a progressive capability enhancement path sequence based on the resource optimization mapping table and the capability assessment result set, and an updated personalized planning report is output.

8. A dynamic multi-dimensional student ability assessment and career planning method according to any one of claims 1 to 7, characterized in that, Based on personalized planning reports, the process of acquiring newly collected student ability data, using a deviation detection model to monitor deviations in student ability improvement progress, and triggering an early warning notification when the deviation exceeds a threshold includes the following steps: Obtain the progressive capability enhancement path sequence from the personalized planning report, and extract the set of stage target capability indicators from the planning path; The system continuously acquires newly collected student ability data through a data acquisition interface to generate a real-time ability dataset. The deviation detection model is invoked to compare the real-time capability dataset with the stage target capability index set and calculate the progress deviation value of each capability dimension. When a progress deviation value in at least one capability dimension exceeds a preset deviation threshold, an early warning event identifier is generated. Based on the warning event identifier, a warning notification push service is triggered to the user terminal.

9. A dynamic, multi-dimensional student ability assessment and career planning system, characterized in that, The system includes: The multi-source data integration and consolidation module is used to collect structured and unstructured student ability data from the education information system, integrate the data, and output the original dataset. The standardized vector generation module is used to clean and quantize the original dataset by calling a standardization algorithm, generate feature vectors containing multi-dimensional capability indicators, and output a standardized feature vector set. The dynamic capability assessment module is used to obtain market demand data through an external data interface, combine the preset professional capability matrix rules and the market demand data, input the standardized feature vector set into a pre-built multi-dimensional capability assessment model for rule matching and weight coefficient update, calculate the capability assessment score and output the capability assessment result set. The visualization analysis module is used to generate a capability distribution radar chart based on the capability assessment result set, and automatically generate an analysis report, outputting visualization charts and analysis reports. The personalized path planning module receives development direction selection instructions from the user interface, applies path planning algorithms to compare ability differences and map resources, generates personalized career planning suggestions and progressive ability improvement paths, and outputs a personalized planning report. The progress monitoring and feedback module is used to acquire newly collected student ability data based on personalized planning reports, call the deviation detection model to monitor deviations in student ability improvement progress, and trigger an early warning notification when the deviation exceeds a threshold.

10. A computer-readable storage medium, characterized in that: The computer program is stored that can be loaded by a processor and executed as described in any one of claims 1 to 8.

Images