Advanced learning path planning method and device based on artificial intelligence, and medium

Abstract

The invention discloses an ascending path planning method and device based on artificial intelligence and a medium, and the method comprises the steps: obtaining the multi-dimensional original data of a student from a plurality of data source ends associated with the same student, and carrying out the preprocessing of the multi-dimensional original data, obtaining multi-dimensional data including structured data and unstructured text data; semantic understanding and analysis are carried out on the multi-dimensional data based on a pre-trained machine learning model, a student portrait is generated, and the student portrait at least comprises a personalized label and an ascending analysis conclusion used for evaluating the ascending potential; according to the student portraits, growth planning paths corresponding to different target school rising paths are generated, and each growth planning path comprises a plurality of stage growth sub-targets from the current stage to the target school rising stage and corresponding action suggestions. According to the invention, automatic generation and prediction of growth planning paths of student portraits and different target school rising paths in the field of low-age education can be realized.

Description

Artificial Intelligence-Based Methods, Devices, and Media for College Admission Path Planning Technical Field

[0001] This invention relates to the field of artificial intelligence technology, and in particular to a method, apparatus and medium for planning academic pathways based on artificial intelligence. Background Technology

[0002] With the increasing demand for personalized education, college planning and development guidance for students have become increasingly important. Currently, educational assessments and college planning services on the market mainly rely on experience-based human consultations or static questionnaire-based psychological or academic assessments. However, experience-based planning is highly dependent on the consultant's personal experience, is subjective, difficult to scale, and costly. On the other hand, static questionnaire-based psychological or academic assessments rely on limited data sources, cannot dynamically and comprehensively reflect students' true characteristics, and lack systematic planning for long-term college strategies and comprehensive quality development. Summary of the Invention

[0003] This invention provides an artificial intelligence-based method, device, and medium for planning college application pathways, in order to solve the technical problems of existing college application planning methods being difficult to scale, costly, and lacking systematic planning for long-term college application strategies and comprehensive quality development.

[0004] To address the aforementioned technical issues, in a first aspect, an artificial intelligence-based method for planning academic pathways is provided, comprising: acquiring multi-dimensional raw data of a student from multiple data sources associated with the same student, and preprocessing the multi-dimensional raw data to obtain multi-dimensional data including structured data and unstructured text data; performing semantic understanding and analysis on the multi-dimensional data based on a pre-trained machine learning model to generate a student profile, wherein the student profile includes at least personalized tags and academic pathway analysis conclusions for assessing academic pathway potential; and generating growth planning paths corresponding to different target academic pathways based on the student profile, wherein each growth planning path includes multiple stage growth sub-goals from the current stage to the target academic pathway stage and corresponding action suggestions.

[0005] Secondly, an artificial intelligence-based college entrance path planning device is provided, including a unit for executing the aforementioned artificial intelligence-based college entrance path planning method.

[0006] Thirdly, a computer-readable storage medium is provided, which stores a computer program that, when executed by a processor, implements the steps of the aforementioned artificial intelligence-based college entrance path planning method.

[0007] In the aforementioned scheme implemented by the AI-based method, device, and medium for planning academic pathways, multi-dimensional raw data is acquired from multiple data sources and preprocessed to obtain multi-dimensional data. A pre-trained machine learning model is used to perform semantic understanding and analysis on the multi-dimensional data, generating a student profile that includes at least personalized tags and academic potential assessment conclusions. Based on the student profile, a growth planning path is generated, corresponding to different target academic pathways, including multiple stage growth sub-goals from the current stage to the target academic stage and corresponding action suggestions. It can be seen that this invention acquires multi-dimensional raw data of students from multiple data sources, dynamically reflecting the students' true characteristics. It enables the automatic generation and prediction of student profiles and growth planning paths for different target academic pathways in early childhood education. Furthermore, the growth planning path includes multiple stage growth sub-goals from the current stage to the target academic stage and corresponding action suggestions, allowing for the parallel generation of multiple customized long-term growth planning paths for a single student, providing clear stage growth sub-goals and corresponding action suggestions, significantly improving the scientific and personalized level of educational decision-making. Attached Figure Description

[0008] Figure 1 is a flowchart illustrating the artificial intelligence-based college entrance path planning method provided in an embodiment of the present invention.

[0009] Figure 2 is a schematic diagram of the specific process of S20 in the AI-based college entrance path planning method shown in Figure 1.

[0010] Figure 3 is a schematic diagram of the specific process of S30 in the AI-based college entrance path planning method shown in Figure 1.

[0011] Figure 4 is a schematic diagram of the specific process of S40 in the AI-based college entrance path planning method shown in Figure 1.

[0012] Figure 5 is a structural block diagram of the artificial intelligence-based college entrance path planning device provided in an embodiment of the present invention. Detailed Implementation

[0013] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0014] It should be understood that, when used in this specification and the appended claims, the terms "comprising" and "including" indicate the presence of the described features, integrals, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or collections thereof.

[0015] It should also be understood that the term "and / or" as used in this specification and the appended claims refers to any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.

[0016] Please refer to Figure 1, which is a schematic flowchart of an AI-based college entrance path planning method provided in an embodiment of the present invention. In the embodiment shown in the figure, the AI-based college entrance path planning method includes the following steps S10-50: S10, obtaining multi-dimensional raw data of the student from multiple data sources associated with the same student, and preprocessing the multi-dimensional raw data to obtain multi-dimensional data including structured data and unstructured text data.

[0017] In this step, raw data on the student's academic performance, learning abilities, personality traits, and family education are collected from various data sources connected to the target student, including teachers, parents, institutions, and the student themselves. Academic performance includes grades (such as Chinese, English, and mathematics) and academic achievements in sports and arts (such as fine arts and music). Learning abilities refer to comprehensive characteristics extracted through standardized cognitive ability assessments, continuous behavioral observation, and specific task performance data, reflecting the student's basic cognitive abilities, learning strategies, and higher-order thinking tendencies. These characteristics may include logical reasoning abilities (such as analyzing and solving problems). The abilities to solve problems and understand relationships between things include: spatial imagination (such as understanding and manipulating spatial figures and relationships), verbal comprehension and expression, learning focus and persistence, planning and monitoring (such as setting learning goals, planning steps, evaluating progress and making adjustments), learning resilience (such as the ability to persevere and adjust when facing difficulties and setbacks), knowledge transfer and integration (such as the ability to apply learned knowledge to new situations and integrate across disciplines), creative thinking and critical thinking, etc. Personality traits may include leadership ability, etc. Family education may include the family education environment and the parents' economic support ability, etc.

[0018] Specifically, the multi-dimensional raw data may include: structured academic data from report cards, multiple-choice answers from standardized tests, text responses from open-ended questionnaires, and teacher comments, etc. In this embodiment, the preprocessing of the multi-dimensional raw data may specifically include: cleaning and normalizing the structured data, segmenting the text data, removing stop words, and vectorizing it using models such as BERT to form a unified multi-dimensional data set.

[0019] Understandably, when using data sources such as teachers', parents', institutions', and students' own devices, information can be entered through web or mobile interfaces. For example, teachers, institutions, or parents can enter structured numerical data such as students' basic information, academic performance, and activity records on their respective source devices. Alternatively, teachers, parents, or students can complete objective question selections on the system to achieve standardized data collection. Or, students can input subjective expressions on their source devices through voice recognition or natural language question answering to obtain corresponding text data.

[0020] S20. Based on a pre-trained machine learning model, perform semantic understanding and analysis on the multi-dimensional data to generate a student profile. The student profile includes at least personalized tags and college entrance analysis conclusions for assessing college entrance potential.

[0021] In this step, a pre-trained machine learning model is used to extract, understand, and analyze features from multi-dimensional data to obtain a student profile that is related to the multi-dimensional data, including at least personalized tags and college entrance analysis conclusions for assessing college entrance potential.

[0022] Specifically, as shown in Figure 2, this step includes S21-S25: ​​S21, Based on a pre-trained machine learning model, feature extraction is performed on the multi-dimensional data to obtain a comprehensive feature vector representing student traits.

[0023] In this step, structured data (such as numerical data like grades) is normalized, and unstructured text data is segmented and features are extracted. Then, a feature fusion network is used to fuse features and generate a comprehensive feature vector. This comprehensive feature vector can represent student characteristics, such as academic performance, learning ability, personality traits, and family education environment.

[0024] S22. Based on the preset talent profile type, the comprehensive feature vector is analyzed and classified using a classification algorithm to determine at least one preset talent profile type to which the student belongs, which serves as the personalized label, and suggestions for further education associated with the preset talent profile type are obtained.

[0025] In this embodiment, the preset talent profile types can be summarized into various talent profile types and their baseline characteristics through training models with a large amount of data. These profiles can also be associated with suitable professional directions and further education suggestions, or with relevant representative figures. Specifically, they can include the following 16 types: All-rounder: A hexagonal talent with outstanding performance in multiple dimensions and cross-disciplinary integration capabilities; associated professional directions include engineering, architectural design, business administration, etc.; further education suggestions are suitable for aiming for top comprehensive universities; the plan should showcase their cross-disciplinary abilities and achievements in engineering (such as AI / robotics), architectural design, business administration, etc.; representative figures can be: Elon Musk, Leonardo da Vinci, and Yuzuru Ito; Influencer: Primarily outstanding in sports, leadership, and artistic achievements; associated with... Specializations include media, international relations, and arts management. Recommended academic paths include top business, media, and design schools. Planning should focus on developing and showcasing cross-cultural communication, public speaking, and experiences that generate social impact through cultural and sporting activities. Notable figures include Oprah Winfrey, Yang Lan, and Howard Schultz. Engineers are typically strong in sports, leadership, and STEM thinking (primarily logical reasoning and spatial imagination). Related specializations include mechanical engineering, computer engineering, and industrial design. Recommended academic paths include top engineering schools. Planning should focus on developing practical skills such as engineering competitions and prototyping, cultivating a hands-on and pragmatic personality. A notable figure is Tony Fadell (the creator of the iPod). Commander: Primarily excels in sports and leadership; related majors include business administration, marketing, and sports management; recommended academic paths include business administration, marketing, and sports management, with a focus on practical experience in team event organization and business planning. Notable figures include Phil Knight (founder of Nike) and Li Ning. Strategist: Primarily excels in leadership, artistic, and STEM thinking; related majors include management science, policy research, and strategic consulting; recommended academic paths include interdisciplinary experimental programs, management science, public policy, and strategic consulting, with a focus on accumulating systematic research, cross-disciplinary project design, and social research experience. Research experience; The Debater: Primarily excels in leadership and humanities / social sciences abilities, with related majors including political science, philosophy, law, and media studies. Further study is recommended to focus on humanities / social sciences, emphasizing critical thinking, debate, and academic writing skills. A representative figure is Socrates. Project Manager: Primarily excels in leadership and STEM thinking, with related majors including data science, industrial engineering, and information systems. Further study is recommended to focus on interdisciplinary fields of STEM and management, such as Industrial Engineering and Operations Research (IEOR), Management Science and Engineering (MS&E), and information systems. Planning should emphasize technical project management, process optimization, and data analysis.Coordinator: Primarily excels in leadership and execution (planning and monitoring abilities), with related majors including Business Administration and Educational Administration. Recommended academic tracks include Educational Administration, Business Administration, and Human Resources. Planning should focus on accumulating reliable experience in team coordination, event management, and administrative support. Innovator: Primarily excels in sports, arts, and STEM thinking, with related majors including Bioengineering and Arts & Technology. Recommended academic tracks include Bioengineering, Arts & Technology, and Integrated Design. Planning should integrate scientific experiments and artistic creation, preparing an innovative portfolio. A representative figure is James Cameron. Applied Scientist: Primarily excels in sports and STEM thinking, with related majors including... This includes Software Engineering, Electrical Engineering, and Applied Mathematics; recommended further studies include Computer Engineering, Software Engineering, Electrical Engineering, and Applied Mathematics, which emphasize practical application of technology. Planning should focus on programming development, technology application projects, and internships. Brain Doctor: Primarily excels in STEM thinking; related majors include Mathematics, Physics, and Computer Science. Recommended further studies include in-depth research paths in STEM fields such as Mathematics, Physics, Chemistry, and Computer Science. Planning should focus on academic competitions, research papers, and long-term academic research. Representative figures include Marie Curie, Richard Feynman, and Charles Kao. Interaction Designer: Primarily excels in both artistic ability and STEM thinking; related majors include Architecture, Interaction Design, and Media Arts. The following are suggested career paths: **Guide:** Suitable for applications to architecture, interaction design, media arts, industrial design, and other majors that integrate technology and aesthetics. A portfolio is required, and participation in research competitions is encouraged. **Mentor:** Primarily strong in sports and arts, with related career paths including arts education, physical education, and community culture. **Admissions:** Suitable for applications to arts education, physical education, psychology, community development, and non-profit management. Career paths should focus on developing practical skills in humanistic care, teaching guidance, and community service. **Artist:** Primarily strong in artistic and humanistic qualities, with related career paths including literature, music, drama, and fine arts. **Admissions:** Suitable for applications to creative majors such as fine arts, literature, music, drama, and film production. The core of the plan lies in preparing a high-quality portfolio with a personal style; Sports stars: primarily excelling in sports, with related professional directions including physical education and competitive sports; recommended academic paths are suitable for students with athletic talents, such as physical education, sports science, and competitive sports. The plan should revolve entirely around specialized sports training, competition results, and sportsmanship. Representative figures include Federer and Kobe Bryant; Operations officers: primarily excelling in leadership and execution, with related professional directions including hotel management, public relations, and supply chain management; recommended academic paths are suitable for applying to hotel management, public relations, supply chain management, and administrative affairs. The plan should enhance practical operational skills in event execution, customer service, and logistical support.

[0026] Specifically, in this step, a classification algorithm is used to calculate the similarity between the comprehensive feature vector and the baseline feature vector of each of the preset talent profile types. Based on the similarity calculation results, the student is classified into one or more of the preset talent profile types whose similarity meets the threshold. In this invention, if the similarity between multiple preset talent profile types and the comprehensive feature vector exceeds the threshold, the student belongs to multiple preset talent profile types simultaneously.

[0027] S23. Guided by the proposed academic direction suggestions, a large language model is invoked to perform semantic understanding on the unstructured text data in the multi-dimensional data, generating a SWOT analysis report that includes the dimensions of strengths, weaknesses, opportunities, and threats.

[0028] In this step, the suggested academic direction associated with the personalized tag is used as the core prompt word and input into the large language model. This guides the model to prioritize the discovery and evaluation of strengths and opportunities that match the characteristics of the personalized tag, as well as potential weaknesses and threats related to the strategic direction, when analyzing individual student data, in order to discover suitable professional directions and interest cultivation for the child.

[0029] Specifically, guided by the proposed college admission directions, the Large Language Model (LLM) is invoked to perform focused semantic understanding of the text data. This involves traversing the text data, extracting keywords or phrases related to the college admission directions (such as "confidence," "procrastination," "likes drawing," "participates in competitions," etc.), and matching and classifying them with a preset semantic feature table to generate a SWOT analysis report that includes the dimensions of strengths, weaknesses, opportunities, and threats. For example, when the suggested academic paths associated with the tag "Project Manager" are input into the large language model, the model receives the instruction: "Based on student data, please analyze their strengths and weaknesses related to 'Project Manager' traits (such as system planning, teamwork, project management, data analysis, etc.), and assess external opportunities and threats related to the 'Industrial Engineering' direction." This generates a targeted SWOT analysis report. For example, their strengths might be "proficient in using Gantt charts to manage group project progress," and their weaknesses might be "passive performance in cross-departmental communication cases." Both are closely related to the personalized tag. Opportunities could include suggestions for in-depth participation in projects or competitions to leverage their strengths, while threats could include conclusions that neglecting communication skills might limit their overall evaluation.

[0030] S24. Based on the structured data in the multi-dimensional data, calculate the college entrance potential index under different target college entrance paths.

[0031] In this embodiment, the target college admission pathways include, but are not limited to, the following five categories: the A undergraduate pathway, which is guided by the National College Entrance Examination (Gaokao), the Strong Foundation Program, independent enrollment, and comprehensive evaluation admission; the B undergraduate pathway, which is guided by university applications including DSE and academic proficiency tests; the C undergraduate pathway, which is guided by the AP curriculum system and the SAT, TOEFL, and Common App activity system; the D undergraduate pathway, which is guided by the A-Level or IB curriculum system and the UCAS application system (including the BTEC arts pathway); and other undergraduate pathways.

[0032] In this step, for each target college entrance path, the preset dimension weights corresponding to the personalized tag and the target college entrance path are called to perform weighted calculations on the structured data of academic performance, learning ability, activity experience and psychological characteristics in the multi-dimensional data to obtain the college entrance potential index of the target college entrance path. In this invention, a mapping relationship can be pre-stored to map different combinations of "personalized tags - target college entrance paths" to a preset dimension weight configuration table. During calculation, the mapping relationship is retrieved based on the current student's personalized tags and target college entrance path combination to obtain the preset dimension weights. Then, the college entrance potential index is calculated by weighting the preset dimension weights of academic performance, learning ability, activity experience, and psychological characteristics for each target college entrance path. For example, for the project manager and C undergraduate path, the preset dimension weights are invoked, and higher weights are given to learning ability (logical reasoning) and activity experience (project practice) under this combination. The college entrance potential index calculated accordingly can more realistically reflect the student's relative competitiveness on this path. For example, the preset dimension weights of academic performance, learning ability, activity experience, and psychological characteristics in the C undergraduate path can be 0.1, 0.4, 0.4, and 0.1, respectively, while academic performance is given a greater weight in the A undergraduate path, and the preset dimension weights of academic performance, learning ability, activity experience, and psychological characteristics (such as learning resilience, emotional fluctuations, etc.) can be 0.6, 0.1, 0.2, and 0.1, respectively.

[0033] S25. Based on the personalized tags, SWOT analysis report, and college entrance potential index, generate a student's fit analysis conclusion under different target college entrance paths, thereby obtaining a student profile containing personalized tags and college entrance analysis conclusions; wherein, the SWOT analysis report, the college entrance potential index, and the fit analysis conclusions together constitute the college entrance analysis conclusions used to assess college entrance potential.

[0034] In this step, based on the student's personalized tags, SWOT analysis report, and the college entrance potential index, combined with the characteristics of students who have successfully entered college under different target college entrance paths pre-stored in the education resource database, the student's suitability analysis conclusion (high, medium-high, or low) can be generated by comparison and prediction. For example, the suitability analysis conclusion "A undergraduate path strong foundation program suitability high, C undergraduate path engineering direction suitability medium, needs to strengthen English and activity background" can be generated.

[0035] For steps S21-S25 above, there are specific suggestions for further education associated with the preset talent profile types. After the preset talent profile types are classified, suggestions for further education can be quickly output to guide students, which greatly improves decision-making efficiency and practicality.

[0036] S30. Based on the student profile, generate growth planning paths corresponding to different target higher education paths. Each growth planning path includes multiple stage growth sub-goals from the current stage to the target higher education stage and corresponding action suggestions.

[0037] Specifically, as shown in Figure 3, this step includes S31-S34: S31, Based on the student profile, identify the gap between the student's current status and the target of each target school admission path.

[0038] S32. Based on the goal and corresponding gap of any of the target education path, break down and set multiple stage growth sub-goals that need to be achieved from the current stage to the target education stage for that target education path.

[0039] In this step, based on the target and the current gap with the target, multiple sub-goals of different stages are generated in chronological order. In this embodiment, the target for the academic advancement path, the gap between the student's current state and the available time (the time period from the current stage to the target academic advancement stage) are used as inputs. Dynamic programming or Monte Carlo tree search algorithms are used to generate the target to be achieved, which is divided into multiple growth sub-goals of different stages.

[0040] S33. For each of the aforementioned stage growth sub-goals, generate implementation strategies including course learning, exam arrangements, and extracurricular activities.

[0041] S34. Transform the implementation strategy into action suggestions with clear time nodes, thereby generating a growth planning path corresponding to the target school admission path, which includes multiple stage growth sub-goals and action suggestions.

[0042] Through the steps S31-S34 above, based on the student profile and target academic path, the system can automatically generate stage-specific growth sub-goals and action suggestions from the current grade to grade 12. For example, if we need to plan the A-level undergraduate path for Xiaoming's Strong Foundation Program, the student's strength lies in mathematical logic while his weakness lies in Chinese / English writing. The goal can be broken down into multiple stages, such as setting stage-specific growth sub-goals for grade 9 (winning a city-level math competition award, improving Chinese scores to above the class average, etc.), stage-specific growth sub-goals for grade 10 (experiencing informatics competitions, etc.), and stage-specific growth sub-goals for grade 11, and corresponding action suggestions (such as the action suggestion for the grade 9 sub-goal being to attend a math competition tutoring class and join a Chinese writing enhancement group, etc.).

[0043] S40. Based on the student profile and the current stage goals in the growth planning path of different target further education paths, intelligently match and generate personalized educational resource recommendations from the educational resource database.

[0044] In this invention, the educational resource database stores various types of data, such as course resources, competition books for different disciplines, social activities, reading lists, research project resources, and portfolio examples. Understandably, each resource item (such as courses, books, and activities) in the educational resource database is converted into a vector through a semantic model and indexed.

[0045] Specifically, as shown in Figure 4, this step includes S41-S42: S41, based on the student profile and the current stage goals in the growth planning paths of different target education paths, semantic retrieval is performed in the education resource database using RAG technology to obtain a set of resources matching the current stage of different target education paths.

[0046] In this step, based on the student profile and the current stage goals of the growth planning path for different target academic paths, the RAG (Retrieval-augmented Generation) technology is used to find the Top-K related resource set in the educational resource database through near nearest neighbor search.

[0047] S42. Based on the matched resource set, generate personalized educational resource recommendations with reasons for recommendation for different target college entrance paths.

[0048] S50. Integrate the student profile, the growth planning path for different target college entrance paths, and the personalized education resource recommendations to generate and output a visual analysis report.

[0049] In this step, a visual analysis report can be generated and output to the parents' end, allowing them to view their child's current status, growth plan path, and action suggestions.

[0050] In some embodiments, the AI-based college entrance pathway planning method may further include: acquiring action feedback information at different stages, and adjusting subsequent stage growth sub-goals based on the action feedback information and the goals of the target college entrance pathway. In this invention, subsequent sub-goals can also be adjusted based on feedback.

[0051] As can be seen from the above scheme, this invention obtains multi-dimensional raw data of students from multiple data sources, which can dynamically reflect the true characteristics of students. It realizes the automatic generation and prediction of student profiles and growth planning paths for different target school entrance paths in the field of early childhood education. Moreover, the growth planning path includes multiple stage growth sub-goals from the current stage to the target school entrance stage and corresponding action suggestions. It can generate multiple customized long-term growth planning paths (such as domestic and overseas school entrance) for a single student in parallel, and provide clear stage growth sub-goals and corresponding action suggestions. It also intelligently matches and generates personalized educational resource recommendations from the educational resource database, which can significantly improve the scientific and personalized level of educational decision-making. It can be seen that this invention adopts the method of current situation gap identification - stage goal decomposition - strategy matching - action transformation - resource recommendation, which can realize the automatic generation and prediction of school entrance paths and growth tasks, and the model has continuous optimization and self-learning capabilities.

[0052] It should be understood that the above method embodiments are described as a series of actions for the sake of simplicity. However, those skilled in the art should know that the present invention is not limited by the described order of actions. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0053] Referring to Figure 5, which is a structural block diagram of the AI-based college entrance path planning device provided in an embodiment of the present invention, the AI-based college entrance path planning device includes an acquisition and processing unit 301, a profile generation unit 302, a path planning unit 303, a resource matching unit 304, a visualization and interaction unit 305, and an optimization and adjustment unit 306. The functional units are described in detail below: The acquisition and processing unit 301 is used to acquire multi-dimensional raw data of a student from multiple data sources associated with the same student, and preprocess the multi-dimensional raw data to obtain multi-dimensional data including structured data and unstructured text data; the profile generation unit 302 is used to perform semantic understanding and analysis on the multi-dimensional data based on a pre-trained machine learning model to generate a student profile, which includes at least personalized tags and college entrance analysis conclusions for assessing college entrance potential; the path planning unit 303 is used to generate growth planning paths corresponding to different target college entrance paths based on the student profile, each growth planning path including... The system comprises multiple growth sub-goals from the current stage to the target higher education stage, along with corresponding action suggestions; a resource matching unit 304, used to intelligently match and generate personalized educational resource recommendations from an educational resource database based on the student profile and the current stage goals in the growth planning paths for different target higher education paths; a visualization interaction unit 305, used to integrate the student profile, the growth planning paths for different target higher education paths, and the personalized educational resource recommendations to generate and output a visualization analysis report; and an optimization and adjustment unit 306, used to obtain action feedback information at different stages and adjust the subsequent stage growth sub-goals based on the action feedback information and the goals of the target higher education path.

[0054] In one embodiment, the profile generation unit 302 is specifically used for: extracting features from the multi-dimensional data based on a pre-trained machine learning model to obtain a comprehensive feature vector representing student traits; analyzing and classifying the comprehensive feature vector according to a preset talent profile type using a classification algorithm to determine at least one preset talent profile type to which the student belongs, as the personalized label, and obtaining suggestions for further education related to the preset talent profile type; using the suggestions for further education as guidance, calling a large language model to perform semantic understanding on the unstructured text data in the multi-dimensional data to generate a SWOT analysis report containing the dimensions of strengths, weaknesses, opportunities, and threats; calculating the potential index for further education under different target paths based on the structured data in the multi-dimensional data; generating a suitability analysis conclusion for the student under different target paths based on the personalized label, the SWOT analysis report, and the potential index, thereby obtaining a student profile containing the personalized label and the potential analysis conclusion; wherein, the SWOT analysis report, the potential index, and the suitability analysis conclusion together constitute the potential analysis conclusion for evaluating further education potential.

[0055] In one embodiment, the portrait generation unit 302 is further configured to: calculate the similarity between the comprehensive feature vector and the baseline feature vector of each of the preset talent portrait types using a classification algorithm; and classify the student into one or more of the preset talent portrait types whose similarity meets the threshold based on the similarity calculation results.

[0056] In one embodiment, the portrait generation unit 302 is further specifically used to: for each target school admission path, call the preset dimension weights corresponding to the personalized tag and the target school admission path, and perform weighted calculations on the structured data in the dimensions of academic performance, learning ability, activity experience and psychological characteristics of the multi-dimensional data to obtain the school admission potential index of the target school admission path.

[0057] In one embodiment, the path planning unit 303 is specifically used for: identifying the gap between the student's current state and the goal of each target higher education path based on the student profile; breaking down and setting multiple stage growth sub-goals that need to be achieved from the current stage to the target higher education stage according to the goal of any target higher education path and the corresponding gap; generating implementation strategies including course learning, exam arrangements, and extracurricular activities for each stage growth sub-goal; and converting the implementation strategies into action suggestions with clear time nodes, thereby generating a growth planning path corresponding to the target higher education path that includes stage growth sub-goals and action suggestions.

[0058] In one embodiment, the resource matching unit 304 is specifically used to: perform semantic retrieval in the educational resource database using RAG technology based on the student profile and the current stage goals in the growth planning paths of different target higher education paths, to obtain a set of resources matched to the current stage of different target higher education paths; and generate personalized educational resource recommendations containing recommendation reasons corresponding to different target higher education paths based on the matched resource sets.

[0059] As can be seen, this invention provides an artificial intelligence-based educational pathway planning device. By acquiring multi-dimensional raw data of students from multiple data sources, it can dynamically reflect the students' true characteristics, realize the automatic generation and prediction of student profiles and growth planning paths for different target educational pathways in the field of early childhood education, and the growth planning path includes multiple stage growth sub-goals from the current stage to the target educational stage and corresponding action suggestions. It also intelligently matches and generates personalized educational resource recommendations from the educational resource database, and constructs a process of current status gap identification - stage goal decomposition - strategy matching - action transformation - resource recommendation. It can realize the automatic generation and prediction of educational pathways and growth tasks, and significantly improve the scientific and personalized level of educational decision-making.

[0060] For specific limitations regarding the AI-based college application path planning device, please refer to the limitations of the AI-based college application path planning method mentioned above, which will not be repeated here. Each unit in the aforementioned AI-based college application path planning device can be implemented entirely or partially through software, hardware, or a combination thereof. These units can be embedded in or independent of the processor in a computer device in hardware form, or stored in the memory of a computer device in software form, so that the processor can call and execute the corresponding operations of each unit.

[0061] In one embodiment, the present invention may also provide a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the various steps of the artificial intelligence-based college entrance path planning method provided in the above embodiments.

[0062] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above division of functional units and modules is only used as an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above.

[0063] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.

Claims

1. A method for planning educational pathways based on artificial intelligence, characterized in that, The AI-based college entrance pathway planning method includes: acquiring multi-dimensional raw data of a student from multiple data sources associated with the same student, and preprocessing the multi-dimensional raw data to obtain multi-dimensional data including structured data and unstructured text data; performing semantic understanding and analysis on the multi-dimensional data based on a pre-trained machine learning model to generate a student profile, the student profile including at least personalized tags and college entrance analysis conclusions for assessing college entrance potential; and generating growth planning paths corresponding to different target college entrance paths based on the student profile, each growth planning path including multiple stage growth sub-goals from the current stage to the target college entrance stage and corresponding action suggestions.

2. The artificial intelligence-based college entrance path planning method as described in claim 1, characterized in that, The pre-trained machine learning model performs semantic understanding and analysis on the multi-dimensional data to generate a student profile. Specifically, this includes: extracting features from the multi-dimensional data using the pre-trained machine learning model to obtain a comprehensive feature vector representing student traits; analyzing and classifying the comprehensive feature vector according to preset talent profile types using a classification algorithm to determine at least one preset talent profile type to which the student belongs, serving as the personalized label, and obtaining suggestions for further education related to the preset talent profile type; using the suggestions for further education as guidance, calling a large language model to perform semantic understanding on the unstructured text data in the multi-dimensional data to generate a SWOT analysis report containing strengths, weaknesses, opportunities, and threats; calculating the further education potential index under different target further education paths based on the structured data in the multi-dimensional data; and generating a suitability analysis conclusion for the student under different target further education paths based on the personalized label, the SWOT analysis report, and the suitability analysis conclusion, thereby obtaining a student profile containing the personalized label and the further education analysis conclusion; wherein the SWOT analysis report, the further education potential index, and the suitability analysis conclusion together constitute the further education analysis conclusion used to assess further education potential.

3. The artificial intelligence-based college entrance path planning method as described in claim 2, characterized in that, The step of analyzing and classifying the comprehensive feature vector according to the preset talent profile type using a classification algorithm specifically includes: calculating the similarity between the comprehensive feature vector and the baseline feature vector of each talent profile type in the preset talent profile type using a classification algorithm; and classifying the student into one or more preset talent profile types whose similarity meets the threshold based on the similarity calculation results.

4. The artificial intelligence-based college entrance path planning method as described in claim 2, characterized in that, The step of calculating the college entrance potential index for different target college entrance paths based on the structured data in the multi-dimensional data specifically includes: for each target college entrance path, calling the preset dimension weights corresponding to the personalized tag and the target college entrance path, and performing weighted calculation on the structured data in the dimensions of academic performance, learning ability, activity experience and psychological characteristics in the multi-dimensional data to obtain the college entrance potential index of the target college entrance path.

5. The artificial intelligence-based college entrance path planning method as described in claim 1, characterized in that, The step of generating growth planning paths corresponding to different target higher education paths based on the student profile specifically includes: identifying the gap between the student's current state and the goal of each target higher education path based on the student profile; breaking down and setting multiple stage growth sub-goals that need to be achieved from the current stage to the target higher education stage according to the goal of any target higher education path and the corresponding gap; generating implementation strategies including course learning, exam arrangements, and extracurricular activities for each stage growth sub-goal; and transforming the implementation strategies into action suggestions with clear time nodes, thereby generating a growth planning path corresponding to the target higher education path that includes multiple stage growth sub-goals and action suggestions.

6. The artificial intelligence-based college entrance path planning method as described in claim 5, characterized in that, The AI-based college entrance path planning method further includes: obtaining action feedback information at different stages, and adjusting the growth sub-goals of subsequent stages based on the action feedback information and the goals of the target college entrance path.

7. The artificial intelligence-based college entrance path planning method as described in claim 1, characterized in that, The AI-based college entrance path planning method further includes: intelligently matching and generating personalized educational resource recommendations from an educational resource database based on the student profile and the current stage goals in the growth planning paths for different target college entrance paths; and integrating the student profile, the growth planning paths for different target college entrance paths, and the personalized educational resource recommendations to generate and output a visual analysis report.

8. The artificial intelligence-based college entrance path planning method as described in claim 7, characterized in that, The process of intelligently matching and generating personalized educational resource recommendations from an educational resource database based on the student profile and the current stage goals in the growth planning paths for different target higher education paths includes: using RAG technology to perform semantic retrieval in the educational resource database according to the student profile and the current stage goals in the growth planning paths for different target higher education paths, obtaining a set of resources matching the current stage for different target higher education paths; and generating personalized educational resource recommendations containing recommendation reasons for different target higher education paths based on the matched resource sets.

9. An artificial intelligence-based college entrance path planning device, characterized in that, Includes a unit for performing the AI-based college admission path planning method as described in any one of claims 1-8.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the steps of the artificial intelligence-based college entrance path planning method as described in any one of claims 1-8.

Images