Employment guidance teaching resource AI generation system

Abstract

The invention discloses an employment guidance teaching resource AI generation system, and relates to the technical field of artificial intelligence, and the system comprises the steps: collecting a target post demand description file, the ability evolution data of a successful entry user, and the current ability data of a student, constructing a standardized post ability vector and ability evolution path graph, extracting the ability vector of the student, extracting a capability difference label set according to the student capability vector and the standardized post capability vector, matching or generating a teaching resource unit according to the capability difference label set, and obtaining a teaching resource unit pushing sequence according to the capability evolution path map, the student capability vector and the standardized post capability vector; the beneficial effects are that the method can dynamically analyze the difference between student ability and post demands, can match personalized teaching resources, and solves the problems of insufficient resource matching precision and lack of ability evaluation in traditional employment guidance, thereby improving the pertinence and timeliness of employment guidance.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence technology, and in particular to an AI generation system for employment guidance teaching resources. Background Art

[0002] Career guidance for college students is an important part of the higher education system. It means providing professional and systematic services to help college students clarify their career direction, improve their employment competitiveness, and ultimately achieve a smooth transition from campus to the workplace.

[0003] Currently, colleges and universities mainly rely on general course arrangements, experience-based lectures and manual consultation methods in the process of providing career guidance. Students usually choose positions and learning paths based on their own interests or vague professional directions. There is a lack of precise guidance on the ability requirements of specific positions. As a result, when students face specific professional positions, they are often unable to accurately identify the gap between their own abilities and job requirements, and it is difficult for them to obtain clearly oriented teaching resource support, resulting in a mismatch between ability development and employment goals.

[0004] Therefore, an AI generation system for employment guidance teaching resources is proposed. Summary of the Invention

[0005] In view of the above-mentioned state of the art, this application is proposed. The embodiments of this application provide an AI-based system for generating career guidance teaching resources, which can dynamically identify the differences between student abilities and job requirements, accurately match personalized teaching resources, and improve the pertinence and timeliness of career guidance.

[0006] According to one aspect of the present application, an AI generation system for employment guidance teaching resources is provided, including: a data acquisition module for collecting a demand description file of a target position, capability evolution data of users who have successfully joined the target position, and current capability data of students to be guided; a position capability parsing module for extracting a set of position capability labels based on the target position description file, and mapping the set of position capability labels to a standardized position capability vector; a success path modeling module for constructing a capability evolution path map based on the capability evolution data, the capability evolution path map including a plurality of time series nodes and corresponding node capability vectors; a student capability portrait generation module for extracting a student capability vector corresponding to the standardized position capability vector structure based on the current capability data of the students to be guided; and a capability difference analysis module for mapping the The dimension items in the difference between the student ability vector and the standardized job ability vector that exceed a preset threshold are mapped into a set of ability difference labels; a resource matching and generation module is used to match the teaching resource units containing learning materials and evaluation materials from the teaching resource database according to the ability difference label set. If the match is not successful, a new teaching resource unit is generated based on the ability difference label; a path matching and push module is used to perform similarity matching on the student ability vector and the standardized job ability vector with the node ability vector in the ability evolution path map, determine the current ability node and the target ability node, and determine the push order of the teaching resource units according to the node order between the current ability node and the target ability node; a resource recommendation module is used to push the teaching resource units to the student-end device according to the push order.

[0007] According to another aspect of the present application, an AI generation method for employment guidance teaching resources is provided, including: collecting a demand description file of a target position, capability evolution data of multiple users who have successfully joined the target position, and current capability data of students to be guided; extracting a set of position capability labels based on the target position description file, and mapping the set of position capability labels into a standardized position capability vector; constructing a capability evolution path map based on the capability evolution data, the capability evolution path map including multiple time series nodes and corresponding node capability vectors; extracting a student capability vector corresponding to the standardized position capability vector structure based on the current capability data of the students to be guided; and comparing the student capability vector with the The dimension items in the difference of the standardized job capability vectors that exceed the preset threshold are mapped into a set of capability difference labels; the teaching resource units containing learning materials and evaluation materials are matched from the teaching resource database according to the capability difference label set. If the match is not successful, a new teaching resource unit is generated based on the capability difference label; the student capability vector and the standardized job capability vector are respectively matched with the node capability vector in the capability evolution path map for similarity, the current capability node and the target capability node are determined, and the push order of the teaching resource units is determined according to the node order between the current capability node and the target capability node; the teaching resource units are pushed to the student terminal device according to the push order.

[0008] Compared with the existing technology, the employment guidance teaching resource AI generation system according to the embodiment of this application can dynamically analyze the differences between students' abilities and job requirements and match personalized teaching resources, thereby solving the problems of insufficient resource matching accuracy and lack of ability assessment in traditional employment guidance, thereby improving the pertinence and timeliness of employment guidance. BRIEF DESCRIPTION OF THE DRAWINGS

[0009] The above and other purposes, features, and advantages of the present application will become more apparent through a more detailed description of the embodiments of the present application in conjunction with the accompanying drawings. The accompanying drawings are intended to provide a further understanding of the embodiments of the present application and constitute a part of the specification. Together with the embodiments of the present application, they are used to explain the present application and do not constitute a limitation of the present application. In the drawings, the same reference numerals generally represent the same components or steps.

[0010] Figure 1 This is a framework diagram of the AI ​​generation system for employment guidance teaching resources of the present invention.

[0011] Figure 2 This is a timing diagram of the AI ​​generation system for employment guidance teaching resources of the present invention.

[0012] Figure 3 This is a flow chart of the AI ​​generation method for employment guidance teaching resources of the present invention. DETAILED DESCRIPTION

[0013] Below, the exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application, and it should be understood that the present application is not limited to the exemplary embodiments described herein.

[0014] Application Overview

[0015] Although universities and vocational training institutions have initially implemented information-based career guidance services, such as providing students with career-related advice through intelligent question-and-answer systems, career assessment tools, or online learning platforms, these methods generally suffer from the following problems:

[0016] 1) It is impossible to structure the job competency requirements, resulting in the lack of job-specificity and competency orientation in employment guidance content.

[0017] 2) Existing learning path recommendation systems mostly rely on course tags or popular learning behaviors, lack empirical modeling based on "successful employment paths", and are unable to provide effective growth guidance.

[0018] 3) Teaching resource recommendations are often based on static rules or course similarity matching, and are unable to dynamically fill or automatically generate gaps in students’ abilities, and are unable to meet the needs of customized, real-time, and systematic ability improvement.

[0019] In response to the above problems, the idea of ​​this application is: taking "job competence" as the core modeling object, establishing a full-process capability driving mechanism from job competence analysis, student competence portrait generation, capability difference identification, resource generation and path recommendation.

[0020] Specifically, this application parses job descriptions through natural language processing and semantic embedding models, extracts standardized job competency labels and generates vector expressions; combines the competency evolution data of historically successful employees to construct a competency growth path map, and realizes the structured expression of typical competency advancement routes; on the student side, extracts current competency data through structured extraction and generates student competency vectors consistent with the job competency dimensions; then analyzes vector differences and evolutionary path node distances to identify the degree of match and gap between students' current competency status and target positions; on this basis, utilizes pre-trained teaching resource generation models and knowledge graph reasoning mechanisms to automatically generate personalized, goal-oriented teaching resource units, and combines the evolutionary order of nodes in the competency evolution path to determine the push path of teaching resources, ultimately realizing an AI generation system for employment guidance teaching resources that is oriented to specific positions, dynamically adapts to competency changes, and supports intelligent content generation and push.

[0021] Exemplary Systems

[0022] Figure 1~Figure 2 The diagram shows an AI generation system for employment guidance teaching resources according to an embodiment of the present application, including: a data collection module, a job capability analysis module, a success path modeling module, a student capability portrait generation module, a capability difference analysis module, a resource matching and generation module, a path matching and push module, and a resource recommendation module.

[0023] Among them, the data acquisition module is used to collect the demand description file of the target position, the ability evolution data of users who have successfully joined the target position, and the current ability data of the students to be guided; the job ability analysis module is used to extract the job ability label set based on the target position description file, and map the job ability label set into a standardized job ability vector; the success path modeling module is used to construct a capability evolution path map based on the ability evolution data, and the capability evolution path map includes multiple time series nodes and corresponding node ability vectors; the student ability portrait generation module is used to extract the student ability vector corresponding to the standardized job ability vector structure according to the current ability data of the students to be guided; the ability difference analysis module is used to compare the student ability vector with the standardized job ability vector structure The dimension items in the force vector difference that exceed the preset threshold are mapped to a set of capability difference labels; the resource matching and generation module is used to match the teaching resource units containing learning materials and evaluation materials from the teaching resource database according to the capability difference label set. If the match is not successful, a new teaching resource unit is generated based on the capability difference label; the path matching and push module is used to perform similarity matching on the student capability vector and the standardized job capability vector with the node capability vector in the capability evolution path map, determine the current capability node and the target capability node, and determine the push order of the teaching resource unit according to the node order between the current capability node and the target capability node; the resource recommendation module is used to push the teaching resource unit to the student terminal device according to the push order.

[0024] Specifically, the system first integrates job requirements and historical success case data to establish a standardized competency assessment system. Then, by comparing the quantitative differences between students' current abilities and the requirements of the target positions, it accurately locates the competency dimensions that need to be improved. The system then matches existing teaching resources based on the competency dimensions that students need to improve. When existing teaching resources cannot meet the needs, new teaching units containing knowledge point explanations and practical tasks are automatically generated. At the same time, referring to the competency development paths of successful people, the system plans a learning sequence for students that conforms to cognitive laws, ensuring that the competency improvement process keeps pace with career development goals.

[0025] Through the above technical solution, this application can detect the matching difference between students' abilities and target positions in real time, automatically generate targeted teaching resources and plan the optimal learning path. This technical solution solves the problems of low resource matching accuracy and lack of scientific basis for learning paths in traditional employment guidance, and improves the fit between students' ability training and career development needs.

[0026] This application further proposes extracting a set of job capability labels based on a target job description file, including: performing word segmentation and entity recognition processing on the target job description file to obtain capability text fragments related to job requirements; classifying and clustering the capability text fragments according to a preset capability classification vocabulary and context semantic model to obtain a set of job capability labels.

[0027] in:

[0028] Word segmentation and entity recognition processing refers to splitting the natural language text in the target job description file into vocabulary units with independent semantics and identifying entity names related to job competencies. This can be achieved using natural language processing tools such as the jieba word segmenter or the spaCy framework to extract key competency elements from unstructured job descriptions.

[0029] A competency classification vocabulary refers to a pre-established structured vocabulary containing industry-wide competency names and their classification relationships. Specifically, it can be constructed using a standard occupational classification system or industry competency framework to provide a standardized classification basis for text fragments.

[0030] A contextual semantic model is a machine learning model that can understand the semantic associations of words in a specific context. Specifically, it can be implemented using pre-trained language models such as BERT or RoBERTa to identify implicit capability-demand associations in text snippets.

[0031] Classification clustering refers to classifying text segments with the same capability attributes according to semantic similarity. It can be implemented using the K-means clustering algorithm or the hierarchical clustering algorithm to eliminate redundant descriptions and generate a standardized set of capability labels.

[0032] Specifically:

[0033] First, after the system receives the target job description file, it breaks the file content into independent vocabulary units through word segmentation. For example, "proficient in Python programming" is broken down into words such as "proficient", "master", "Python", and "programming".

[0034] Then, identify entities related to capabilities, for example, identify "Python programming" as a technical capability entity;

[0035] Next, the extracted competency text fragments are fed into the contextual semantic model and combined with standard terms in the competency classification vocabulary. For example, "Python programming" is mapped to the "Python development skills" label under the "Programming language competency" category.

[0036] Finally, a clustering algorithm is used to merge semantically similar segments. For example, "proficient in data analysis tools" and "proficient in using Excel for data modeling" are clustered into the "data analysis ability" label, thereby generating a structured set of job ability labels.

[0037] Through the above technical solution, the system can identify and classify the common characteristics of job description documents with differentiated expressions, thereby providing a unified benchmark for subsequent capability vector construction and ensuring the accuracy of teaching resource matching.

[0038] This application further proposes mapping a set of job capability labels into a standardized job capability vector, including: vector encoding the set of job capability labels through a preset semantic embedding algorithm; normalizing and dimensionally aligning the vector encoding results according to a preset job capability dimension structure to obtain a standardized job capability vector.

[0039] in:

[0040] The preset semantic embedding algorithm refers to a natural language processing technology that converts text labels into numerical vectors. Specifically, this can be achieved using a word embedding model or a sentence embedding model. For example, Word2Vec, BERT, or similar algorithms can be used to convert the semantic information of the job competency label into a numerical representation in a high-dimensional vector space.

[0041] The preset job competency dimension structure refers to a pre-defined standardized competency indicator system, which can be implemented using an industry-wide competency framework or job competency model. For example, competencies can be divided into fixed dimensions such as professional skills, communication skills, and project management skills to ensure that the competency vectors of different positions or students are comparable.

[0042] Through the above technical solution, this application can convert unstructured job competency labels into standardized vectors with unified dimensions and semantic structures, providing a reliable data basis for subsequent analysis of student competency differences.

[0043] This application further proposes to construct a capability evolution path map based on capability evolution data, including: time-series annotation of capability evolution data, and extraction of node capability vectors corresponding to time-series nodes; stage-clustering of node capability vectors according to a preset state transition modeling algorithm to obtain multiple capability evolution paths; and constructing a capability evolution path map containing node capability vectors and their directed edge relationships based on the time sequence and capability transformation relationship between capability evolution paths.

[0044] in:

[0045] Time series annotation refers to the time series marking of the capability status at different time points in the capability evolution data. It can be implemented by timestamp annotation or event interval annotation to determine the order of capability changes.

[0046] A node capability vector is a vectorized representation of a user's capability status at a specific time point. Specifically, this can be achieved by encoding the capability description text using a semantic embedding algorithm to quantify the capability development trajectory.

[0047] State transition modeling algorithms refer to computational models used to identify stage changes in capability evolution. Specifically, they can be implemented using hidden Markov models or dynamic time warping algorithms to discover key turning points in capability development.

[0048] Stage clustering refers to classifying nodes with similar capability change patterns into the same evolutionary path. This can be achieved using hierarchical clustering or spectral clustering algorithms to form multiple typical capability development branches.

[0049] The capability evolution path map refers to the dynamic process of capability development displayed in the form of a graph structure. Specifically, it can be implemented using a directed graph data structure to visualize the relationship between different capability improvement paths.

[0050] Specifically, when constructing a capability evolution path map, the historical capability data of users before they join the target position is first divided into time series. For example, each month's skill assessment report is used as an independent time series node. Each node is converted into a numerical representation including dimensions such as professional skills and communication skills through vectorization processing. Subsequently, the ability change amplitude between adjacent nodes is analyzed through the state transition modeling algorithm. When a significant jump in the capability dimension is detected, it is marked as a stage dividing point. Based on these dividing points, the capability evolution process is divided into several development stages, and user data with the same stage division pattern are clustered into similar paths. The final map not only contains the capability vector of each node, but also uses directed edges to mark the conversion relationship between different stages. For example, the edge from the primary skill node to the intermediate skill node represents the direction of capability improvement.

[0051] Through the above technical solution, this application can convert discrete capability development data into a structured path map, so that the subsequent system can automatically generate a learning path plan that conforms to the laws of capability development based on the connection relationship of the directed edges in the map, avoiding the cognitive load problem caused by the stacking of learning resources in traditional methods.

[0052] This application further proposes that the extraction of student capability vectors corresponding to the standardized job capability vector structure includes: uniformly formatting the student's current capability data and extracting the capability information fragment text of the student to be guided; constructing capability feature mapping rules based on the job capability label set, and extracting features of the capability information fragments according to the feature mapping rules to obtain an initial student capability vector that matches the job capability label; normalizing and dimensionally reorganizing the initial student capability vector to obtain a student capability vector corresponding to the standardized job capability vector structure.

[0053] in:

[0054] Unified formatting refers to converting student ability data from different sources or formats into a unified data structure. This can be achieved using data cleaning tools and standardized templates to eliminate the interference of data format differences on subsequent analysis.

[0055] The ability feature mapping rule refers to the correspondence between text features and vector dimensions established based on the set of job ability labels. Specifically, it can be implemented using a keyword matching algorithm and a semantic similarity model to ensure that student ability information fragments are accurately mapped to standardized dimensions.

[0056] Normalization processing refers to converting ability characteristics of different dimensions or numerical ranges into a unified numerical interval. Specifically, it can be achieved by using maximum and minimum value normalization or Z-score normalization algorithm to eliminate the magnitude differences between different ability dimensions.

[0057] Specifically, the student's current ability data is first input into the data cleaning tool, and redundant information is removed and the ability information fragment text is extracted through standardized templates. Then, based on the preset set of job ability labels, a keyword matching algorithm is used to identify feature items related to job requirements from the ability information fragments, and the feature items are mapped to the corresponding vector dimensions through the semantic similarity model to generate an initial student ability vector. After the vector is processed by the normalization algorithm, it is reorganized according to the dimensional structure of the standardized job ability vector, and finally a student ability vector that is completely consistent with the job ability vector structure is generated. Therefore, the difference analysis between student abilities and job requirements can be carried out under the same dimensional structure, avoiding ability assessment deviations caused by inconsistent vector structures.

[0058] Through the above technical solution, this application solves the problem of ability assessment errors caused by the inconsistency between student ability data and job ability vector structure, ensures the dimensional alignment and numerical comparability of ability difference analysis, and provides a structurally consistent data foundation for subsequent precise matching of teaching resources.

[0059] This application further proposes that dimension reorganization includes: dimensional rearrangement and missing zero filling of the normalized student ability vector according to the dimensional structure of the standardized job ability vector to obtain a student ability vector that is consistent with the number of dimensions and semantic structure of the standardized job ability vector.

[0060] in:

[0061] Dimension rearrangement refers to rearranging the dimensional items of the student competency vector according to the dimensional order of the standardized job competency vector. This can be achieved by using a matrix transposition operation or an index mapping algorithm to eliminate the dimensional order differences generated by different data sources.

[0062] Missing zero filling processing means that when there is a missing dimension in the dimensional structure corresponding to the standardized job ability vector of the student ability vector, a zero value is filled in the missing dimension position. Specifically, the missing dimension position can be identified through the dimension alignment detection algorithm to maintain the consistency of the number of vector dimensions.

[0063] Specifically, when generating student ability vectors, since the ability data sources of different students may have different dimensional orders or some dimensions are missing, the dimensional order of the student ability vector is forced to be aligned to the dimensional structure defined by the standardized job ability vector through dimensional rearrangement operation, and the missing dimensional items are filled through zero-padding operation, so as to ensure that the two vectors are numerically compared in the same dimensional space during the subsequent ability difference analysis.

[0064] For example, when the standardized job competency vector includes three dimensions: programming ability, communication ability, and project management ability, if a student's ability data only includes two dimensions: programming ability and communication ability, then zero is added to the project management ability dimension.

[0065] Through the above technical solution, this application eliminates the dimensional structure differences between student ability vectors and job ability vectors, ensures the dimensional consistency of ability difference analysis, avoids ability assessment deviations caused by dimensional misalignment or missingness, and provides a reliable data basis for subsequent precise matching of teaching resources.

[0066] This application further proposes mapping the dimension items in the difference between the student ability vector and the standardized job ability vector that exceed a preset threshold into a set of ability difference labels, including: calculating the dimension difference between the standardized job ability vector and the student ability vector dimension by dimension to obtain an ability difference vector; screening the dimension items in the ability difference vector that are higher than the preset threshold as insufficient ability dimensions; and generating an ability difference label set based on the mapping relationship between the insufficient ability dimensions and the job ability label set.

[0067] in:

[0068] The ability gap vector is a numerical vector formed by calculating the difference between the standardized job ability vector and the student ability vector dimension by dimension. Specifically, it can be implemented by vector subtraction operation, which is used to quantify the actual gap between students in each ability dimension.

[0069] The capability difference label set refers to the set formed by mapping the capability deficiency dimension to the corresponding job capability label. Specifically, it can be implemented by using a label mapping table, that is, searching the preset dimension-label correspondence table and converting the numerical capability deficiency dimension into an interpretable text label.

[0070] Specifically, when calculating the ability difference vector, the standardized job ability vector and the student ability vector are first aligned under the same dimensional structure, and then numerical subtraction operations are performed on each dimension.

[0071] For example, if the value of the standardized job competency vector in the "data analysis ability" dimension is 0.85, and the value of the student competency vector in this dimension is 0.52, the corresponding competency difference is 0.33. When the preset threshold is 0.3, the dimension is judged as a dimension of insufficient competency. Subsequently, by querying the mapping relationship between the pre-established dimensions and job competency labels, for example, mapping the "data analysis ability" dimension to the "Python data processing" label, a set of competency difference labels containing descriptions of specific competency deficiencies is finally generated.

[0072] Through the above technical solution, this application can accurately identify students' specific deficiencies in specific job competency dimensions and generate structured competency difference labels, providing a clear directional basis for subsequent teaching resource matching.

[0073] This application further proposes generating new teaching resource units, including: generating task objectives that match ability difference labels through a pre-trained teaching resource generation model; retrieving knowledge points, skill requirements, and evaluation requirements related to the task objectives through a preset teaching knowledge graph to obtain a draft teaching resource structure including the basic composition framework and elements of the teaching resources; generating teaching resource text content including learning materials and evaluation materials based on the teaching resource structure draft through a preset natural language generation model; and structurally encapsulating the teaching resource text content to obtain a teaching resource unit.

[0074] in:

[0075] A pre-trained teaching resource generation model refers to an artificial intelligence model that has been trained with a large amount of teaching resource data. Specifically, it can be implemented using a generative model based on the Transformer architecture, which is used to automatically generate corresponding learning task objectives based on ability difference labels.

[0076] The preset teaching knowledge graph refers to a structured knowledge base that contains the relationship between knowledge points and skill levels. It can be implemented by using a knowledge network built based on a graph database to provide a standardized content framework for teaching resources.

[0077] A natural language generation model is an algorithmic model that can convert structured data into natural language text. Specifically, it can be implemented using a sequence generation model based on LSTM or GPT, and is used to convert teaching resource elements into highly readable learning materials.

[0078] Structured encapsulation refers to the process of organizing and storing text content in a standard format. Specifically, data can be encapsulated in XML or JSON format to ensure the callability of generated resources in the system.

[0079] Specifically, when the system detects that the existing teaching resource library lacks teaching content with corresponding ability difference labels, it first analyzes the learning objectives corresponding to the ability difference labels through the teaching resource generation model. For example, for the label of "insufficient ability to apply data analysis tools", it generates the task goal of "mastering Python data processing tools". Then, through the teaching knowledge graph, it retrieves the knowledge points related to the task goal, such as core skill requirements such as Pandas library operations and data visualization methods, to form a teaching framework consisting of three parts: theoretical explanation, practical exercises, and results evaluation. Next, the natural language generation model automatically generates supporting learning material texts based on the framework, such as operation guides containing code examples, and evaluation materials containing test questions. Finally, the generated text content is standardized and packaged according to the chapter structure to form a teaching resource unit that can be directly pushed.

[0080] Through the above technical solution, this application achieves the dynamic generation of teaching resource units that precisely match ability differences, solving the problem of disconnection between teaching resources and job competency requirements in existing career guidance. The knowledge graph ensures the systematic construction of teaching resources, avoiding the knowledge blind spots that exist in traditional manually compiled resources. At the same time, the automated generation process improves the efficiency of resource updates, ensuring that students can access targeted learning content in a timely manner.

[0081] This application further proposes that determining the push order of teaching resource units includes: extracting a directed path from the current capability node to the target capability node based on the capability evolution path map; arranging the teaching resource units according to the numerical change order of the node capability vectors of each intermediate node in the directed path on the dimension corresponding to the capability difference label, to obtain the push order of the teaching resource units.

[0082] in:

[0083] A directed path refers to the ability development route from the starting node to the end node formed by the connection relationship between nodes in the ability evolution path map. Specifically, it can be achieved by using a graph traversal algorithm to extract the shortest path or optimal path from the current ability node to the target ability node. It is used to describe the staged trajectory of students' ability improvement.

[0084] The order of numerical change refers to the increasing or decreasing trend of the value of the node capability vector in a specific dimension over time. Specifically, it can be achieved by using the time series analysis method to perform trend fitting on the node capability vector to reflect the dynamic change law in the capability difference dimension.

[0085] Specifically, when determining the order of push, first search for an effective path connecting the student's current ability node and the target ability node corresponding to the target position in the ability evolution path map. For example, when a student needs to improve the programming ability dimension, the system will filter out the path branches where the value of this dimension shows a step-by-step increase. Then, by analyzing the magnitude of the change in the values ​​of the adjacent nodes in the path on the dimensions corresponding to the ability difference labels, the teaching resource units are sorted according to the gradient of ability improvement. For example, if the value of the data structure ability dimension of an intermediate node in the path is increased by 30% compared with the previous node, the corresponding data structure teaching resource unit will be pushed first.

[0086] Through the above technical solution, this application can generate a teaching resource push sequence that conforms to the law of ability development based on the dynamic differences between students' current ability status and the requirements of the target position. This solution solves the problem of mismatch between the resource push sequence and the actual ability improvement path in traditional methods, allowing students to obtain teaching content that matches their current ability gaps at each learning stage, thereby improving the pertinence of employment guidance and the scientific nature of path planning.

[0087] Exemplary Methods

[0088] Figure 3The figure shows a flow chart of the AI ​​generation method of employment guidance teaching resources according to an embodiment of the present application, including: collecting the demand description file of the target position, the ability evolution data of multiple users who have successfully entered the target position, and the current ability data of the students to be guided; extracting a set of job ability labels based on the target position description file, and mapping the set of job ability labels to a standardized job ability vector; constructing a capability evolution path map based on the ability evolution data, the capability evolution path map including multiple time series nodes and corresponding node ability vectors; extracting a student ability vector corresponding to the standardized job ability vector structure based on the current ability data of the students to be guided; and mapping the student ability vector to the standardized job ability vector structure. The dimension items in the difference of the standardized job competency vectors that exceed the preset threshold are mapped into a set of competency difference labels; the teaching resource units containing learning materials and assessment materials are matched from the teaching resource database according to the set of competency difference labels. If the match is not successful, a new teaching resource unit is generated based on the competency difference label; the student competency vector and the standardized job competency vector are respectively matched with the node competency vector in the competency evolution path map for similarity, the current competency node and the target competency node are determined, and the push order of the teaching resource units is determined according to the node order between the current competency node and the target competency node; the teaching resource units are pushed to the student-end device in the push order.

[0089] In one example, extracting a set of job capability labels based on a target job description file includes: performing word segmentation and entity recognition processing on the target job description file to obtain capability text fragments related to job requirements; classifying and clustering the capability text fragments according to a preset capability classification vocabulary and contextual semantic model to obtain a set of job capability labels.

[0090] In one example, mapping a set of job capability labels into a standardized job capability vector includes: vector encoding the set of job capability labels through a preset semantic embedding algorithm; normalizing and dimensionally aligning the vector encoding results according to a preset job capability dimension structure to obtain a standardized job capability vector.

[0091] In one example, constructing a capability evolution path map based on capability evolution data includes: time-series annotation of capability evolution data and extracting node capability vectors corresponding to time-series nodes; performing stage clustering on node capability vectors according to a preset state transition modeling algorithm to obtain multiple capability evolution paths; and constructing a capability evolution path map containing node capability vectors and their directed edge relationships based on the time sequence and capability transformation relationship between capability evolution paths.

[0092] In one example, extracting a student capability vector corresponding to a standardized job capability vector structure includes: uniformly formatting the student's current capability data and extracting the capability information fragment text of the student to be guided; constructing a capability feature mapping rule based on a set of job capability labels, and performing feature extraction on the capability information fragment according to the feature mapping rule to obtain an initial student capability vector that matches the job capability label; normalizing and dimensionally reorganizing the initial student capability vector to obtain a student capability vector corresponding to the standardized job capability vector structure.

[0093] In one example, dimension reorganization includes: dimensional rearrangement and missing zero filling of the normalized student competency vector according to the dimensional structure of the standardized job competency vector, to obtain a student competency vector that is consistent with the standardized job competency vector in terms of dimensional number and semantic structure.

[0094] In one example, mapping the dimension items in the difference between the student ability vector and the standardized job ability vector that exceed a preset threshold into a set of ability difference labels includes: calculating the dimension difference between the standardized job ability vector and the student ability vector dimension by dimension to obtain an ability difference vector; screening the dimension items in the ability difference vector that are higher than the preset threshold as insufficient ability dimensions; and generating an ability difference label set based on the mapping relationship between the insufficient ability dimensions and the job ability label set.

[0095] In one example, generating a new teaching resource unit includes: generating task objectives that match ability difference labels through a pre-trained teaching resource generation model; retrieving knowledge points, skill requirements, and evaluation requirements related to the task objectives through a preset teaching knowledge graph to obtain a teaching resource structure draft that includes the basic composition framework and elements of the teaching resources; generating teaching resource text content including learning materials and evaluation materials based on the teaching resource structure draft through a preset natural language generation model; and structurally encapsulating the teaching resource text content to obtain a teaching resource unit.

[0096] In one example, determining the push order of teaching resource units includes: extracting a directed path from the current capability node to the target capability node based on the capability evolution path map; arranging the teaching resource units according to the numerical change order of the node capability vectors of each intermediate node in the directed path on the dimension corresponding to the capability difference label, to obtain the push order of the teaching resource units.

[0097] The above description of the disclosed aspects is provided to enable any person skilled in the art to make or use the present application. Various modifications to these aspects will be readily apparent to those skilled in the art, and the general principles defined herein may be applied to other aspects without departing from the scope of the present application. Therefore, the present application is not intended to be limited to the aspects shown herein, but rather to be accorded the widest scope consistent with the principles and novel features disclosed herein.

[0098] The above description has been provided for the purpose of illustration and description. Furthermore, this description is not intended to limit the embodiments of the present application to the forms disclosed herein. Although a number of example aspects and embodiments have been discussed above, those skilled in the art will recognize certain variations, modifications, alterations, additions, and sub-combinations thereof.

Claims

1. The AI ​​generation system for employment guidance teaching resources is characterized by: include: A data collection module is used to collect the requirements description document of the target position, the ability evolution data of multiple users who have successfully joined the target position, and the current ability data of the students to be mentored; A job capability parsing module, configured to extract a job capability tag set based on the target job description file and map the job capability tag set into a standardized job capability vector; A success path modeling module, configured to construct a capability evolution path map based on the capability evolution data, wherein the capability evolution path map includes a plurality of time series nodes and corresponding node capability vectors; A student capability profile generation module is used to extract a student capability vector corresponding to the standardized job capability vector structure based on the current capability data of the student to be guided; A capability difference analysis module, configured to map dimension items exceeding a preset threshold in the difference between the student capability vector and the standardized job capability vector into a capability difference label set; a resource matching and generation module, configured to match teaching resource units containing learning materials and assessment materials from a teaching resource database according to the ability difference tag set, and if no match is successful, generate new teaching resource units based on the ability difference tags; A path matching and push module is used to perform similarity matching on the student capability vector and the standardized job capability vector with the node capability vector in the capability evolution path map, determine the current capability node and the target capability node, and determine the push order of the teaching resource units according to the node sequence between the current capability node and the target capability node; The resource recommendation module is used to push the teaching resource units to the student terminal device according to the push order.

2. The AI ​​generation system for employment guidance teaching resources according to claim 1 is characterized in that: Extracting a job capability tag set based on the target job description file includes: Perform word segmentation and entity recognition on the target job description file to obtain capability text segments related to job requirements; The capability text segments are classified and clustered according to a preset capability classification vocabulary and a context semantic model to obtain the position capability label set.

3. The AI ​​generation system for employment guidance teaching resources according to claim 2 is characterized in that: Mapping the job capability label set to a standardized job capability vector includes: Performing vector encoding on the job capability label set by using a preset semantic embedding algorithm; The vector encoding result is normalized and dimensionally aligned according to a preset job capability dimension structure to obtain the standardized job capability vector.

4. The employment guidance teaching resource AI generation system according to claim 1, characterized in that: The constructing of a capability evolution path map based on the capability evolution data includes: Performing time series annotation on the capability evolution data and extracting node capability vectors corresponding to time series nodes; Performing stage clustering on the node capability vectors according to a preset state transition modeling algorithm to obtain multiple capability evolution paths; According to the time sequence and capability conversion relationship between the capability evolution paths, a capability evolution path graph including node capability vectors and directed edge relationships is constructed.

5. The AI ​​generation system for career guidance teaching resources according to claim 1 is characterized in that: The extracting of the student capability vector corresponding to the standardized job capability vector structure includes: Performing unified formatting on the student's current ability data and extracting a text fragment of the ability information of the student to be guided; Constructing an ability feature mapping rule based on the job ability label set, and performing feature extraction on the ability information fragment according to the feature mapping rule to obtain an initial student ability vector that matches the job ability label; The initial student competence vector is normalized and dimensionally reorganized to obtain a student competence vector corresponding to the standardized job competence vector structure.

6. The employment guidance teaching resource AI generation system according to claim 5 is characterized in that: The dimensional reorganization includes: performing dimensional rearrangement and missing zero filling processing on the normalized student ability vector according to the dimensional structure of the standardized job ability vector, so as to obtain a student ability vector that is consistent with the number of dimensions and semantic structure of the standardized job ability vector.

7. The AI ​​generation system for career guidance teaching resources according to claim 1 is characterized in that: Mapping the dimension items exceeding a preset threshold in the difference between the student ability vector and the standardized job ability vector into an ability difference label set includes: Calculate the dimension-by-dimension difference between the standardized job competency vector and the student competency vector to obtain a competency difference vector; Filtering dimension items in the capability difference vector that are higher than a preset threshold as capability deficiency dimensions; The capability difference label set is generated according to the mapping relationship between the capability deficiency dimension and the position capability label set.

8. The AI ​​generation system for career guidance teaching resources according to claim 1 is characterized in that: The generating of a new teaching resource unit comprises: Generate a task objective that matches the ability difference label through a pre-trained teaching resource generation model; Retrieve knowledge points, skill requirements, and assessment requirements related to the task objectives through a preset teaching knowledge graph to obtain a draft teaching resource structure including the basic framework and elements of the teaching resource; Generate teaching resource text content including learning materials and assessment materials based on the teaching resource structure draft through a preset natural language generation model; The teaching resource text content is structurally packaged to obtain the teaching resource unit.

9. The AI ​​generation system for career guidance teaching resources according to claim 1, characterized in that: Determining the order of pushing the teaching resource units includes: Extracting a directed path from a current capability node to a target capability node according to the capability evolution path graph; The teaching resource units are arranged in the order of numerical changes of the node capability vectors of the intermediate nodes in the directed path on the dimension corresponding to the capability difference label to obtain the push order of the teaching resource units.

10. An AI-generated method for employment guidance teaching resources, using the system according to any one of claims 1 to 9, characterized in that: include: Collect the target position's requirement description document, the capability evolution data of multiple users who have successfully entered the target position, and the current capability data of the students to be mentored; Extracting a job capability tag set based on the target job description file, and mapping the job capability tag set into a standardized job capability vector; Constructing a capability evolution path map based on the capability evolution data, wherein the capability evolution path map includes a plurality of time series nodes and corresponding node capability vectors; Extracting a student capability vector corresponding to the standardized job capability vector structure based on the current capability data of the student to be guided; Mapping dimension items exceeding a preset threshold in the difference between the student ability vector and the standardized job ability vector into an ability difference label set; Matching teaching resource units containing learning materials and assessment materials from a teaching resource database according to the ability difference tag set, and if no match is successful, generating new teaching resource units based on the ability difference tags; Performing similarity matching between the student capability vector and the standardized job capability vector and the node capability vector in the capability evolution path map, determining the current capability node and the target capability node, and determining the order of pushing the teaching resource units according to the node sequence between the current capability node and the target capability node; The teaching resource units are pushed to the student terminal devices according to the pushing order.

Images