Engineering bidding and tendering knowledge dynamic pushing method and system based on student behavior analysis

Abstract

The invention discloses an engineering bidding knowledge dynamic pushing method and system based on student behavior analysis, and belongs to the technical field of online education. Through full-dimensional behavior data acquisition and accurate demand feature mining, multi-dimensional behavior data of student knowledge demands are realized, a stereoscopic and accurate demand portrait is generated, current weak links and preferences of students are clarified, potential demands can be pre-judged, and it is ensured that pushed contents are highly matched with the student demands; based on demand feature core elements and multi-dimensional matching conditions, an intelligent matching engine is built to carry out bidirectional comparison and comprehensive evaluation, secondary verification and adaptation index screening are combined, target knowledge quality is ensured, pushing modes and opportunities are selected in combination with learning habits of students, core knowledge points and reinforcing contents are preferentially pushed, and pushing frequency is dynamically controlled. According to the method, blindness and interference of traditional pushing are avoided, the knowledge viewing rate and utilization efficiency are improved, and pushing service better meets individual requirements of trainees through optimization and continuous iteration.

Description

Technical Field

[0001] This invention relates to the field of online education technology, and in particular to a method and system for dynamically pushing engineering bidding knowledge based on student behavior analysis. Background Technology

[0002] Current learning methods for engineering bidding knowledge largely rely on traditional offline training, fixed course videos, or static documents, which have significant limitations. Firstly, knowledge delivery is often standardized and uniform, failing to adequately consider individual learners' knowledge bases, learning progress, and weaknesses. This results in low relevance between the delivered content and learners' actual needs, leading to low learning efficiency for some. Secondly, existing learning models lack in-depth analysis of learners' learning behaviors, making it difficult to accurately capture learners' knowledge preferences and potential needs, and hindering dynamic and adaptive knowledge delivery. Furthermore, engineering bidding knowledge resources are scattered and outdated, making it difficult and time-consuming for learners to independently select effective knowledge, further impacting learning outcomes. Summary of the Invention

[0003] The purpose of this invention is to provide a method and system for dynamically pushing engineering bidding knowledge based on student behavior analysis, so as to solve the problems mentioned in the background art.

[0004] To achieve the above objectives, the present invention provides the following technical solution: a method for dynamically pushing engineering bidding knowledge based on student behavior analysis, comprising the following steps:

[0005] The system collects real-time, comprehensive behavioral data of students during their learning of engineering bidding knowledge through student learning terminals. This behavioral data includes data on course browsing, video learning, answering exercises, knowledge point retrieval, online interaction, and learning duration.

[0006] Based on behavioral data, the core needs of learners are mined through behavioral feature extraction and demand analysis algorithms to generate a knowledge demand profile. The core needs include knowledge preferences, weaknesses, learning progress and learning frequency.

[0007] Construct a multi-dimensional engineering bidding knowledge resource base that includes policies and regulations, case studies, theoretical knowledge, practical guidance, and simulation exercises, and add keyword tags and difficulty level indicators to the knowledge resources;

[0008] Based on the knowledge needs of students, matching and intelligent filtering are performed in the knowledge resource base to determine the appropriate target knowledge content.

[0009] By combining students' learning habits and historical learning time distribution, the push method and timing are selected to dynamically push the target knowledge to the students' terminals.

[0010] Furthermore, the learning terminals collect real-time, comprehensive behavioral data on students during their learning process of engineering bidding knowledge, including:

[0011] Real-time capture of students' operational behaviors on the learning platform is used as raw behavioral data. These operational behaviors include course click counts, video playback progress, pause frequency, fast forward frequency, exercise answer records, wrong question submission content, knowledge point search keywords, online question and discussion records, learning time distribution, and the duration of a single learning session.

[0012] Add timestamps and student identification information to the collected raw behavioral data, preprocess the raw behavioral data and temporarily store it according to data type.

[0013] Furthermore, based on behavioral data, behavioral feature extraction and needs analysis algorithms are used to uncover the core needs of trainees, specifically including:

[0014] Extract key feature indicators from preprocessed behavioral data;

[0015] Among them, knowledge preference characteristics are determined by the percentage of course clicks, the distribution of video learning time, and the frequency of knowledge point retrieval; weak link characteristics are derived by analyzing the error rate of exercises, the distribution of knowledge points in wrong questions, and the time spent learning key and difficult content; learning progress characteristics are determined based on the comparison between the preset learning plan and the actual learning completion, including the percentage of completed courses, the amount of unlearned content, and the learning progress deviation value; learning frequency characteristics are calculated by the number of times of learning per unit time, the frequency of learning in fixed time periods, and the interval between knowledge reviews.

[0016] A multi-dimensional needs assessment system is constructed, and key characteristic indicators are weighted and integrated to form a complete profile of students' knowledge needs.

[0017] Furthermore, a multi-dimensional knowledge resource base for engineering bidding and tendering will be constructed, encompassing policies and regulations, case studies, theoretical knowledge, practical guidance, and simulated exercises. Specifically, this includes:

[0018] Collect knowledge resources in the field of engineering bidding, including relevant national and local policies and regulations, typical bidding case analysis reports, professional theoretical textbooks, practical process guidance videos, and simulation exercise sets of different difficulty levels;

[0019] According to the engineering bidding process, knowledge resources are divided into different core knowledge modules, which include bidding planning, bidding document preparation, bid document preparation, bid opening and evaluation, contract signing and performance management, and specific knowledge sub-items under each core knowledge module.

[0020] Add multi-dimensional keyword tags to each knowledge resource. The keyword tags include knowledge topic, core test points and applicable scenarios. Mark the difficulty level according to the content complexity of the knowledge resource. The difficulty level includes basic, intermediate and expert levels.

[0021] Build a knowledge resource indexing system to associate and store the storage address, tag information, difficulty level, and content summary of knowledge resources.

[0022] Furthermore, based on the learners' knowledge needs, matching and intelligent filtering are performed in the knowledge resource base to determine suitable target knowledge content, specifically including:

[0023] The knowledge needs characteristics of trainees are obtained and the knowledge needs characteristics are analyzed in depth to determine the core elements of the knowledge needs characteristics. The knowledge needs characteristics are at least one set, and the core elements include knowledge topic preferences, knowledge weaknesses, learning progress stages, and difficulty level.

[0024] The core elements are prioritized and analyzed to determine the weight of each element in the matching and filtering process, and a knowledge matching task is performed based on the weight of each element.

[0025] When performing the knowledge matching task, the corresponding knowledge screening criteria are determined based on the specific requirements of the core elements, and multi-dimensional matching conditions are constructed based on the screening criteria. The matching conditions include at least topic relevance, difficulty suitability, progress fit, and reinforcement targeting.

[0026] A knowledge intelligent matching engine is built based on the matching conditions. At the same time, the attribute tags of each knowledge content in the knowledge resource library are obtained, and the association link between demand features and knowledge content is constructed through the knowledge intelligent matching engine based on the attribute tags.

[0027] Based on the aforementioned association links, the core elements and matching conditions are compared bidirectionally with the knowledge content in the knowledge resource base in the knowledge intelligent matching engine. At the same time, a comprehensive adaptation evaluation is performed by combining the weight ratio of each element, and candidate knowledge content is selected based on the evaluation results through the aforementioned association links.

[0028] A temporary resource pool is constructed based on the candidate knowledge content, and the knowledge content in the temporary resource pool is subjected to secondary verification. The verification dimensions include the timeliness of the content, the accuracy of the information, and the adaptability of the presentation format. The verification results are fed back to the knowledge intelligent matching engine.

[0029] Based on the feedback results, the final filtering rules for candidate knowledge content are determined in the knowledge intelligent matching engine, and the candidate knowledge content is accurately filtered and sorted based on the filtering rules.

[0030] Based on the filtering and sorting results, the adaptation index of the target knowledge content is determined, and when the adaptation index is greater than or equal to the preset adaptation threshold, the determination of the target knowledge content is completed.

[0031] Furthermore, based on the evaluation results, candidate knowledge content is selected through the aforementioned association links, specifically including:

[0032] The process involves obtaining the core element requirements corresponding to the knowledge demand characteristics, and analyzing the degree of fit of each knowledge content with any one of the core element requirements based on these requirements. Specific steps include:

[0033] Compare the attribute tags of the knowledge content with the core element requirements one by one to determine the degree of fit for each requirement, including three levels: complete fit, partial fit, and no fit.

[0034] By combining the weighting of each core element, the level of fit of the knowledge content across all requirements is comprehensively evaluated to form a comprehensive fit score.

[0035] The overall adaptation score is compared with a preset score threshold.

[0036] If the overall fit score is less than the preset score threshold, it is determined that the knowledge content does not meet the student's knowledge needs and is excluded from the candidate range.

[0037] Otherwise, the knowledge content is determined to meet the student's knowledge needs and is included in the candidate knowledge content set for a second verification stage.

[0038] Furthermore, based on students' learning habits and historical learning time distribution, the method and timing of push notifications are selected, specifically including:

[0039] By analyzing students' historical learning data, the preferred push methods for students are identified, including pop-up reminders, message notifications, pinning of the learning page, email push, and APP push.

[0040] By analyzing the historical learning time distribution data of students, the time periods when students have the highest learning frequency and the best learning state are identified as the best time to push notifications.

[0041] Adjust the push strategy according to different types of target knowledge content, and prioritize pushing core knowledge points and content to strengthen weak areas;

[0042] Set up a push frequency control mechanism to dynamically adjust the number of pushes based on the student's learning intensity;

[0043] When pushing target knowledge, include a brief introduction to the knowledge, learning suggestions, and a description of its relevance to the learners' needs.

[0044] Furthermore, the engineering bidding knowledge dynamic push system based on student behavior analysis, applied to the aforementioned engineering bidding knowledge dynamic push method based on student behavior analysis, includes:

[0045] The behavioral data collection module is deployed on the student learning terminal to collect students' full-dimensional behavioral data in real time during the learning process of engineering bidding knowledge, generate behavioral datasets, and support the synchronous collection and classification of multiple types of data.

[0046] The demand feature mining module is used to mine the knowledge demand features of students based on behavioral datasets and to build a student demand feature profile through behavioral feature extraction and demand analysis algorithms.

[0047] The knowledge resource repository module is used to store multi-dimensional knowledge resources in the field of engineering bidding, add tags and difficulty level indicators to knowledge resources, and provide dynamic updates and fast search functions for knowledge resources.

[0048] The intelligent matching and filtering module is used to accurately match and intelligently filter the knowledge resource base based on the student's needs profile to determine the appropriate target knowledge content.

[0049] The dynamic push module is used to combine students' learning habits and historical learning time distribution to select the appropriate push method and the best push time to dynamically push target knowledge to students' terminals.

[0050] Furthermore, the knowledge resource base module includes:

[0051] The resource collection unit is used to collect policy and regulation information, case studies, theoretical knowledge, practical guidance, and simulation exercises in the field of engineering bidding.

[0052] Resource classification units are used to divide knowledge resources into modules and sub-items according to the engineering bidding process, and to build a hierarchical knowledge resource system.

[0053] The tag level labeling unit is used to add multi-dimensional keyword tags and difficulty level indicators to each knowledge resource to form standardized resource attribute information;

[0054] The dynamic update unit is used to periodically collect the latest knowledge resources and delete outdated content;

[0055] The indexing and retrieval unit is used to build a knowledge resource indexing system based on resource attribute information.

[0056] Furthermore, it also includes a feedback optimization module, which is used to obtain feedback data from students on the target knowledge being pushed, and adjust the parameters and weight allocation scheme of the knowledge demand feature analysis model based on the feedback data.

[0057] Compared with the prior art, the beneficial effects of the present invention are:

[0058] 1. This invention achieves in-depth insight into students' knowledge needs through comprehensive behavioral data collection and precise demand feature mining. It collects multi-dimensional behavioral data such as course browsing, exercise answering, and knowledge point retrieval, and combines dynamic collection frequency adjustment to ensure data comprehensiveness and smooth terminal operation. It uses professional algorithms to extract core demand indicators and construct a multi-dimensional evaluation system to generate a three-dimensional and accurate demand profile, clarifying students' current weaknesses and specific preferences, predicting potential needs, ensuring that the pushed content is highly consistent with students' needs, and improving the relevance of learning.

[0059] 2. This invention relies on diversified resource collection and the construction of a standardized knowledge resource base. Resources are collected through multiple channels such as government websites, industry associations, academic institutions, and user contributions, covering various types of content including policies and regulations, case studies, and practical guidance. This ensures the authority and comprehensiveness of the resources. Through standardized processing such as classification, tagging, and dynamic updates, a knowledge system with clear hierarchy and convenient retrieval is constructed. This ensures that the knowledge content keeps up with changes in industry policies and practices, enabling rapid and accurate resource retrieval. It provides rich and high-quality content support for knowledge matching and meets the learning needs of students in different scenarios.

[0060] 3. This invention achieves accurate and efficient delivery of knowledge content through intelligent matching and personalized dynamic push mechanisms. Based on the core elements of demand characteristics and multi-dimensional matching conditions, an intelligent matching engine is built for two-way comparison and comprehensive evaluation. Combined with secondary verification and adaptation index filtering, the quality of target knowledge is ensured. The push method and timing are selected based on students' learning habits, prioritizing the push of core knowledge points and supplementary content, and dynamically controlling the push frequency. This model avoids the blindness and interference of traditional pushes, improves knowledge viewing rate and utilization efficiency, and continuously iterates through a feedback optimization mechanism, making the push service more tailored to students' personalized needs, significantly improving the learning experience and knowledge mastery. Attached Figure Description

[0061] Fig. 1 This is a schematic diagram of the dynamic push method for engineering bidding knowledge according to the present invention;

[0062] Fig. 2 This is a schematic diagram of the target knowledge content matching and filtering process of the present invention. Detailed Implementation

[0063] 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 embodiments of the present invention, and not all embodiments. 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.

[0064] Please see Figs. 1-2 The present invention provides the following technical solutions:

[0065] The method for dynamically pushing engineering bidding knowledge based on student behavior analysis includes the following steps:

[0066] The system collects real-time, multi-dimensional behavioral data of students during their learning process of engineering bidding knowledge through student learning terminals. This behavioral data includes data on course browsing, video learning, answering exercises, knowledge point retrieval, online interaction, and learning duration.

[0067] Based on behavioral data, through behavioral feature extraction and demand analysis algorithms, we can mine the core demand features of learners and generate a knowledge demand feature profile. The core demand features include knowledge preferences, weak points, learning progress and learning frequency.

[0068] Construct a multi-dimensional engineering bidding knowledge resource base that includes policies and regulations, case studies, theoretical knowledge, practical guidance, and simulation exercises, and add keyword tags and difficulty level indicators to the knowledge resources;

[0069] Based on the knowledge needs of students, matching and intelligent filtering are performed in the knowledge resource base to determine the appropriate target knowledge content.

[0070] By combining students' learning habits and historical learning time distribution, the push method and timing are selected to dynamically push the target knowledge to the students' terminals.

[0071] The system collects real-time, multi-dimensional behavioral data on learners during their learning process of engineering bidding knowledge through student learning terminals, specifically including:

[0072] Real-time capture of students' operational behaviors on the learning platform serves as raw behavioral data. These behaviors include course click counts, video playback progress, pause frequency, fast-forward frequency, exercise answer records, submission of incorrect answers, keywords for knowledge point searches, online question and discussion records, distribution of learning time periods, and duration of each learning session.

[0073] Add timestamps and student identification information to the collected raw behavioral data, preprocess the raw behavioral data and temporarily store it according to data type;

[0074] A dynamic data acquisition frequency adjustment mechanism is set up to adaptively adjust the acquisition interval according to the student's learning status, thereby reducing terminal resource consumption while ensuring data integrity.

[0075] In the above embodiments, real-time collection of multi-dimensional behavioral data through student learning terminals enables comprehensive and accurate capture of the student's learning process. This includes not only basic behaviors such as course browsing and video learning, but also in-depth behavioral data such as details of answering exercises, keywords for knowledge point searches, and online interaction records. This ensures the comprehensiveness and richness of the behavioral data. Adding timestamps and student identification information provides a unique identifier for subsequent data correlation analysis, avoiding data confusion. By dynamically adjusting the collection frequency, the resource consumption of student terminals is effectively reduced while ensuring data integrity, improving terminal operation smoothness, and preventing data collection from interfering with the student's learning experience.

[0076] Based on behavioral data, and through behavioral feature extraction and needs analysis algorithms, we uncover the core needs of learners, specifically including:

[0077] Extract key feature indicators from preprocessed behavioral data;

[0078] Among them, knowledge preference characteristics are determined by the percentage of course clicks, the distribution of video learning time, and the frequency of knowledge point retrieval; weak link characteristics are derived by analyzing the error rate of exercises, the distribution of knowledge points in wrong questions, and the time spent learning key and difficult content; learning progress characteristics are determined based on the comparison between the preset learning plan and the actual learning completion, including the percentage of completed courses, the amount of unlearned content, and the learning progress deviation value; learning frequency characteristics are calculated by the number of times of learning per unit time, the frequency of learning in fixed time periods, and the interval between knowledge reviews.

[0079] A multi-dimensional needs assessment system is constructed, and key characteristic indicators are weighted and integrated to form a complete profile of students' knowledge needs.

[0080] In the above embodiments, based on preprocessed behavioral data, professional algorithms are used to mine core demand features, realizing the transformation from massive behavioral data to precise demand profiles. Through the extraction of multi-dimensional key feature indicators, student needs are deconstructed from four core dimensions: knowledge preference, weaknesses, learning progress, and learning frequency, breaking the limitations of traditional single-dimensional demand judgment. A multi-dimensional demand feature evaluation system is constructed, and each indicator is weighted and integrated, highlighting key demand features while maintaining comprehensiveness, resulting in a more three-dimensional and accurate profile of student knowledge needs. This approach can deeply mine students' potential knowledge needs, not only clarifying their current weaknesses and preferences but also predicting the knowledge content they may need in their subsequent learning process. This provides a scientific basis for accurate matching and delivery of subsequent knowledge, ensuring that the delivered content highly matches the student's needs and improving learning effectiveness.

[0081] Construct a multi-dimensional engineering bidding knowledge resource base that includes policies and regulations, case studies, theoretical knowledge, practical guidance, and simulation exercises, specifically including:

[0082] Collect knowledge resources in the field of engineering bidding, including relevant national and local policies and regulations, typical bidding case analysis reports, professional theoretical textbooks, practical process guidance videos, and simulation exercise sets of different difficulty levels;

[0083] Among these measures, the latest bidding and tendering policies, regulations, industry standards, and management methods are collected from the official websites of national and local government authorities to ensure the authority and timeliness of policy-related resources.

[0084] Collaborate with engineering bidding industry associations and professional academic institutions to obtain certified typical case analysis reports, industry research results and academic papers to enrich case and theoretical resources;

[0085] Integrate the core content of officially published engineering bidding professional textbooks, tutoring books and practical manuals, extract structured knowledge points and form theoretical knowledge modules;

[0086] We collect practical teaching videos, online courses, and live sharing content from senior experts and front-line practitioners in the industry, and select highly practical practical guidance resources.

[0087] By collaborating with engineering bidding training institutions and vocational education platforms, we can acquire simulated exercise sets, specialized training question banks, and assessment papers that have been verified through teaching practice, thereby improving our exercise resources.

[0088] It includes publicly available bidding announcements, winning bid notices, and project performance cases published by major domestic bidding and tendering platforms, extracting knowledge elements from real business scenarios;

[0089] Establish a user contribution mechanism to encourage students and industry practitioners to upload compliant and high-quality knowledge resources, which will be included in the knowledge resource library after professional review and approval, thus broadening the channels for resource collection.

[0090] According to the engineering bidding process, knowledge resources are divided into different core knowledge modules. The core knowledge modules include bidding planning, bidding document preparation, bid document preparation, bid opening and evaluation, contract signing and performance management. Each core knowledge module is further subdivided into specific knowledge sub-items.

[0091] Add multi-dimensional keyword tags to each knowledge resource. Keyword tags include knowledge topic, core test points and applicable scenarios. Mark the difficulty level according to the content complexity of the knowledge resource. The difficulty level includes basic, intermediate and expert levels.

[0092] Establish a dynamic update mechanism for knowledge resources, regularly include the latest policies and regulations, new bidding cases and industry practical skills, and delete outdated and invalid knowledge content to ensure the timeliness and practicality of the knowledge resource base;

[0093] Build a knowledge resource indexing system that links the storage address, tag information, difficulty level, and content summary of knowledge resources for rapid retrieval and matching.

[0094] In the above embodiments, a multi-dimensional engineering bidding knowledge resource base is constructed. Through diversified collection channels and standardized processing procedures, the authority, comprehensiveness, and practicality of the knowledge resources are ensured. By classifying by business process, adding multi-dimensional tags, and marking difficulty levels, a hierarchical and easily searchable knowledge system is constructed. With the help of a dynamic update mechanism and indexing system, it is ensured that the knowledge resources can keep up with changes in industry policies and practices in a timely manner, and that rapid retrieval and accurate matching of knowledge resources are achieved, providing rich and high-quality resource support for efficient knowledge dissemination in the future.

[0095] Based on the learners' knowledge needs, matching and intelligent filtering are performed in the knowledge resource base to determine suitable target knowledge content, specifically including:

[0096] The knowledge needs characteristics of learners are obtained and analyzed in depth to determine the core elements of the knowledge needs characteristics. Among them, the knowledge needs characteristics are at least one set, and the core elements include knowledge topic preferences, knowledge weaknesses, learning progress stages, and difficulty level.

[0097] Priority ranking analysis is performed on the core elements to determine the weight of each element in the matching and filtering process, and knowledge matching tasks are performed based on the weight of each element.

[0098] When performing knowledge matching tasks, the corresponding knowledge screening criteria are determined based on the specific requirements of the core elements, and multi-dimensional matching conditions are constructed based on the screening criteria. The matching conditions include at least topic relevance, difficulty suitability, progress fit, and reinforcement targeting.

[0099] A knowledge intelligent matching engine is built based on matching conditions. At the same time, the attribute tags of each knowledge content in the knowledge resource library are obtained, and the association link between demand features and knowledge content is constructed based on the attribute tags through the knowledge intelligent matching engine.

[0100] Based on the association links, the knowledge intelligent matching engine compares the core elements and matching conditions with the knowledge content in the knowledge resource base in two directions. At the same time, it performs a comprehensive adaptation evaluation by combining the weight ratio of each element, and selects candidate knowledge content based on the evaluation results through the association links.

[0101] A temporary resource pool is constructed based on candidate knowledge content, and the knowledge content in the temporary resource pool is verified a second time. The verification dimensions include the timeliness of the content, the accuracy of the information, and the adaptability of the presentation format. The results of the verification are fed back to the knowledge intelligent matching engine.

[0102] Based on the feedback results, the final filtering rules for candidate knowledge content are determined in the knowledge intelligent matching engine, and the candidate knowledge content is accurately filtered and sorted based on the filtering rules.

[0103] The matching index of the target knowledge content is determined based on the screening and sorting results, and the determination of the target knowledge content is completed when the matching index is greater than or equal to the preset matching threshold.

[0104] Among these, candidate knowledge content is selected based on the evaluation results and through correlation links, specifically including:

[0105] The process involves identifying the core element requirements corresponding to the characteristics of knowledge needs, and analyzing the degree to which each knowledge item fits any one of these core element requirements. Specific steps include:

[0106] Compare the attribute tags of the knowledge content with the core element requirements one by one to determine the degree of fit for each requirement, including three levels: complete fit, partial fit, and no fit.

[0107] By combining the weighting of each core element, the level of fit of the knowledge content across all requirements is comprehensively evaluated to form a comprehensive fit score.

[0108] Compare the overall fit score with the preset score threshold;

[0109] If the overall fit score is less than the preset score threshold, the knowledge content is determined not to meet the learner's knowledge needs and is excluded from the candidate range;

[0110] Otherwise, the knowledge content is determined to meet the student's knowledge needs and is included in the candidate knowledge content set for a second verification stage.

[0111] In the above embodiments, by deeply analyzing the characteristics of knowledge needs, constructing multi-dimensional matching conditions and an intelligent matching engine, the precise matching of knowledge content and learner needs is achieved. Priority ranking and weighting analysis of core elements are introduced to highlight the impact of key needs on the matching results, improving the targeting of the matching. A knowledge intelligent matching engine and related links are built to achieve two-way comparison and comprehensive adaptation evaluation of demand characteristics and knowledge content, ensuring the comprehensiveness and accuracy of the matching results. A temporary resource pool is constructed for secondary verification, further filtering from dimensions such as timeliness, accuracy, and presentation format adaptability to eliminate unsuitable knowledge content, ensuring the quality of the target knowledge content. The final target knowledge is determined based on the adaptation index, and a preset adaptation threshold is set as a screening standard to avoid adaptation bias caused by subjective judgment, ensuring that the pushed knowledge content can accurately meet the core and potential needs of learners, improving the effectiveness of knowledge push and learner learning efficiency.

[0112] Based on students' learning habits and historical learning time distribution, the method and timing of push notifications are selected, specifically including:

[0113] By analyzing students' historical learning data, we can identify students' preferred push notification methods, including pop-up reminders, message notifications, pinning of learning pages, email pushes, and app pushes. Students can choose and modify their own push notification methods.

[0114] By analyzing the historical learning time distribution data of students, we can determine the time periods when students have the highest learning frequency and the best learning state as the best time to push notifications, thus avoiding ineffective pushes during non-learning periods.

[0115] Adjust the push strategy according to different types of target knowledge content, prioritize pushing core knowledge points and content to strengthen weak areas, and push extended knowledge in a timely manner according to the student's learning progress.

[0116] Set up a push frequency control mechanism to dynamically adjust the number of pushes based on the student's learning intensity, so as to avoid excessive pushes that may disturb the student.

[0117] When pushing target knowledge, include a brief introduction to the knowledge, learning suggestions, and explanations of its relevance to the learners' needs to help them quickly understand the value of the knowledge and the key learning points.

[0118] In the above embodiments, the push method and timing are selected by combining students' learning habits and historical time distribution, realizing the personalization and humanization of knowledge push. By analyzing historical data, the push method preferred by students is identified and can be modified independently, fully respecting students' usage habits and personalized needs, improving students' acceptance of the push service, and accurately locating the time when students have the highest learning frequency and best state as the push time, avoiding interference caused by push during non-learning periods, improving the viewing rate and utilization efficiency of knowledge content, adjusting the push strategy for different types of knowledge content, prioritizing the push of core knowledge points and supplementary content, ensuring that students can obtain key knowledge first, helping to achieve breakthroughs in key areas and improve weak areas, setting up a push frequency control mechanism to dynamically adjust the number of pushes, avoiding excessive pushes that may cause student resentment, and attaching knowledge introductions and learning suggestions with pushes to help students quickly grasp the value of knowledge and learning key points, improving the pertinence and efficiency of learning, and further optimizing the personalized learning experience.

[0119] The engineering bidding knowledge dynamic push system based on student behavior analysis, applied to the aforementioned engineering bidding knowledge dynamic push method based on student behavior analysis, includes:

[0120] The behavioral data collection module is deployed on the student learning terminal to collect students' full-dimensional behavioral data in real time during the learning process of engineering bidding knowledge, generate behavioral datasets, and support the synchronous collection and classification of multiple types of data.

[0121] The demand feature mining module is used to mine the knowledge demand features of students based on behavioral datasets and to build a student demand feature profile through behavioral feature extraction and demand analysis algorithms.

[0122] The knowledge resource repository module is used to store multi-dimensional knowledge resources in the field of engineering bidding, add tags and difficulty level indicators to knowledge resources, and provide dynamic updates and fast search functions for knowledge resources.

[0123] The intelligent matching and filtering module is used to accurately match and intelligently filter the knowledge resource base based on the student's needs profile to determine the appropriate target knowledge content.

[0124] The dynamic push module is used to combine students' learning habits and historical learning time distribution to select the appropriate push method and the best push time to dynamically push the target knowledge to the student's terminal.

[0125] The feedback optimization module is used to obtain feedback data from students on the target knowledge being pushed, and to adjust the parameters and weight allocation scheme of the knowledge demand feature analysis model based on the feedback data.

[0126] The knowledge resource base module includes:

[0127] The resource collection unit is used to collect policy and regulation information, case studies, theoretical knowledge, practical guidance, and simulation exercises in the field of engineering bidding.

[0128] Resource classification units are used to divide knowledge resources into modules and sub-items according to the engineering bidding process, and to build a hierarchical knowledge resource system.

[0129] The tag level labeling unit is used to add multi-dimensional keyword tags and difficulty level indicators to each knowledge resource to form standardized resource attribute information;

[0130] The dynamic update unit is used to periodically collect the latest knowledge resources and delete outdated content;

[0131] The indexing and retrieval unit is used to build a knowledge resource indexing system based on resource attribute information.

[0132] In the above embodiments, the engineering bidding knowledge dynamic push system based on student behavior analysis achieves efficient and coordinated operation of various functions through modular design, improving the overall system's stability and scalability. The behavior data collection module ensures comprehensive data input, the demand feature mining module transforms data into demands, the knowledge resource base module provides rich and high-quality resource support, the intelligent matching and filtering module completes precise matching, the dynamic push module achieves efficient delivery, and the feedback optimization module ensures continuous system iteration. The various units under the knowledge resource base module further refine their functions, from resource collection and classification to tagging, dynamic updates, and index construction, ensuring standardized and efficient knowledge resource management. This not only meets the current core needs of dynamic knowledge push but also reserves space for future functional expansion, allowing for flexible addition of new modules or optimization of existing functions based on industry development and changes in student needs.

[0133] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A method for dynamically pushing engineering bidding knowledge based on student behavior analysis, characterized in that, Includes the following steps: The system collects real-time, comprehensive behavioral data of students during their learning of engineering bidding knowledge through student learning terminals. This behavioral data includes data on course browsing, video learning, answering exercises, knowledge point retrieval, online interaction, and learning duration. Based on behavioral data, the core needs of learners are mined through behavioral feature extraction and demand analysis algorithms to generate a knowledge demand profile. The core needs include knowledge preferences, weaknesses, learning progress and learning frequency. Construct a multi-dimensional engineering bidding knowledge resource base that includes policies and regulations, case studies, theoretical knowledge, practical guidance, and simulation exercises, and add keyword tags and difficulty level indicators to the knowledge resources; Based on the knowledge needs of students, matching and intelligent filtering are performed in the knowledge resource base to determine the appropriate target knowledge content. By combining students' learning habits and historical learning time distribution, the push method and timing are selected to dynamically push the target knowledge to the students' terminals.

2. The method for dynamically pushing engineering bidding knowledge based on student behavior analysis as described in claim 1, characterized in that, The system collects real-time, multi-dimensional behavioral data of trainees during their learning process of engineering bidding knowledge through their learning terminals, specifically including: Real-time capture of students' operational behaviors on the learning platform is used as raw behavioral data. These operational behaviors include course click counts, video playback progress, pause frequency, fast forward frequency, exercise answer records, wrong question submission content, knowledge point search keywords, online question and discussion records, learning time distribution, and the duration of a single learning session. Add timestamps and student identification information to the collected raw behavioral data, preprocess the raw behavioral data and temporarily store it according to data type.

3. The method for dynamically pushing engineering bidding knowledge based on student behavior analysis as described in claim 1, characterized in that, Based on behavioral data, and through behavioral feature extraction and needs analysis algorithms, we uncover the core needs of learners, specifically including: Extract key feature indicators from preprocessed behavioral data; Among them, knowledge preference characteristics are determined by the percentage of course clicks, the distribution of video learning time, and the frequency of knowledge point retrieval; weak link characteristics are derived by analyzing the error rate of exercises, the distribution of knowledge points in wrong questions, and the time spent learning key and difficult content; learning progress characteristics are determined based on the comparison between the preset learning plan and the actual learning completion, including the percentage of completed courses, the amount of unlearned content, and the learning progress deviation value; learning frequency characteristics are calculated by the number of times of learning per unit time, the frequency of learning in fixed time periods, and the interval between knowledge reviews. A multi-dimensional needs assessment system is constructed, and key characteristic indicators are weighted and integrated to form a complete profile of students' knowledge needs.

4. The method for dynamically pushing engineering bidding knowledge based on student behavior analysis as described in claim 1, characterized in that, Construct a multi-dimensional engineering bidding knowledge resource base that includes policies and regulations, case studies, theoretical knowledge, practical guidance, and simulation exercises, specifically including: Collect knowledge resources in the field of engineering bidding, including relevant national and local policies and regulations, typical bidding case analysis reports, professional theoretical textbooks, practical process guidance videos, and simulation exercise sets of different difficulty levels; According to the engineering bidding process, knowledge resources are divided into different core knowledge modules, which include bidding planning, bidding document preparation, bid document preparation, bid opening and evaluation, contract signing and performance management, and specific knowledge sub-items under each core knowledge module. Add multi-dimensional keyword tags to each knowledge resource. The keyword tags include knowledge topic, core test points and applicable scenarios. Mark the difficulty level according to the content complexity of the knowledge resource. The difficulty level includes basic, intermediate and expert levels. Build a knowledge resource indexing system to associate and store the storage address, tag information, difficulty level, and content summary of knowledge resources.

5. The method for dynamically pushing engineering bidding knowledge based on student behavior analysis as described in claim 1, characterized in that, Based on the learners' knowledge needs, matching and intelligent filtering are performed in the knowledge resource base to determine suitable target knowledge content, specifically including: The knowledge needs characteristics of trainees are obtained and the knowledge needs characteristics are analyzed in depth to determine the core elements of the knowledge needs characteristics. The knowledge needs characteristics are at least one set, and the core elements include knowledge topic preferences, knowledge weaknesses, learning progress stages, and difficulty level. The core elements are prioritized and analyzed to determine the weight of each element in the matching and filtering process, and a knowledge matching task is performed based on the weight of each element. When performing the knowledge matching task, the corresponding knowledge screening criteria are determined based on the specific requirements of the core elements, and multi-dimensional matching conditions are constructed based on the screening criteria. The matching conditions include at least topic relevance, difficulty suitability, progress fit, and reinforcement targeting. A knowledge intelligent matching engine is built based on the matching conditions. At the same time, the attribute tags of each knowledge content in the knowledge resource library are obtained, and the association link between demand features and knowledge content is constructed through the knowledge intelligent matching engine based on the attribute tags. Based on the aforementioned association links, the core elements and matching conditions are compared bidirectionally with the knowledge content in the knowledge resource base in the knowledge intelligent matching engine. At the same time, a comprehensive adaptation evaluation is performed by combining the weight ratio of each element, and candidate knowledge content is selected through the association links based on the evaluation results. A temporary resource pool is constructed based on the candidate knowledge content, and the knowledge content in the temporary resource pool is verified a second time. The verification dimensions include the timeliness of the content, the accuracy of the information, and the adaptability of the presentation format. The results of the verification are fed back to the knowledge intelligent matching engine. Based on the feedback results, the final filtering rules for candidate knowledge content are determined in the knowledge intelligent matching engine, and the candidate knowledge content is accurately filtered and sorted based on the filtering rules. Based on the filtering and sorting results, the adaptation index of the target knowledge content is determined, and when the adaptation index is greater than or equal to the preset adaptation threshold, the determination of the target knowledge content is completed.

6. The method for dynamically pushing engineering bidding knowledge based on student behavior analysis as described in claim 5, characterized in that, Based on the evaluation results, candidate knowledge content is selected through the aforementioned association links, specifically including: The process involves obtaining the core element requirements corresponding to the knowledge demand characteristics, and analyzing the degree of fit of each knowledge content with any one of the core element requirements based on these requirements. Specific steps include: Compare the attribute tags of the knowledge content with the core element requirements one by one to determine the degree of fit for each requirement, including three levels: complete fit, partial fit, and no fit. By combining the weighting of each core element, the level of fit of the knowledge content across all requirements is comprehensively evaluated to form a comprehensive fit score. The overall adaptation score is compared with a preset score threshold. If the overall fit score is less than the preset score threshold, it is determined that the knowledge content does not meet the student's knowledge needs and is excluded from the candidate range. Otherwise, the knowledge content is determined to meet the student's knowledge needs and is included in the candidate knowledge content set for a second verification stage.

7. The method for dynamically pushing engineering bidding knowledge based on student behavior analysis as described in claim 1, characterized in that, Based on students' learning habits and historical learning time distribution, the method and timing of push notifications are selected, specifically including: By analyzing students' historical learning data, the preferred push methods for students are identified, including pop-up reminders, message notifications, pinning of the learning page, email push, and APP push. By analyzing the historical learning time distribution data of students, the time periods when students have the highest learning frequency and the best learning state are identified as the best time to push notifications. Adjust the push strategy according to different types of target knowledge content, and prioritize pushing core knowledge points and content to strengthen weak areas; Set up a push frequency control mechanism to dynamically adjust the number of pushes based on the student's learning intensity; When pushing target knowledge, include a brief introduction to the knowledge, learning suggestions, and a description of its relevance to the learners' needs.

8. A dynamic knowledge delivery system for engineering bidding based on learner behavior analysis, applied in the dynamic knowledge delivery method for engineering bidding based on learner behavior analysis as described in any one of claims 1-7, characterized in that, include: The behavioral data collection module is deployed on the student learning terminal to collect students' full-dimensional behavioral data in real time during the learning process of engineering bidding knowledge, generate behavioral datasets, and support the synchronous collection and classification of multiple types of data. The demand feature mining module is used to mine the knowledge demand features of students based on behavioral datasets and to build a student demand feature profile through behavioral feature extraction and demand analysis algorithms. The knowledge resource repository module is used to store multi-dimensional knowledge resources in the field of engineering bidding, add tags and difficulty level indicators to knowledge resources, and provide dynamic updates and fast search functions for knowledge resources. The intelligent matching and filtering module is used to accurately match and intelligently filter the knowledge resource base based on the student's needs profile to determine the appropriate target knowledge content. The dynamic push module is used to combine students' learning habits and historical learning time distribution to select the appropriate push method and the best push time to dynamically push target knowledge to students' terminals.

9. The dynamic knowledge push system for engineering bidding based on student behavior analysis as described in claim 8, characterized in that, The knowledge resource base module includes: The resource collection unit is used to collect policy and regulation information, case studies, theoretical knowledge, practical guidance, and simulation exercises in the field of engineering bidding. Resource classification units are used to divide knowledge resources into modules and sub-items according to the engineering bidding process, and to build a hierarchical knowledge resource system. The tag level labeling unit is used to add multi-dimensional keyword tags and difficulty level indicators to each knowledge resource to form standardized resource attribute information; The dynamic update unit is used to periodically collect the latest knowledge resources and delete outdated content; The indexing and retrieval unit is used to build a knowledge resource indexing system based on resource attribute information.

10. The dynamic knowledge push system for engineering bidding based on student behavior analysis as described in claim 8, characterized in that, It also includes a feedback optimization module, which is used to obtain feedback data from students on the target knowledge being pushed, and adjust the parameters and weight allocation scheme of the knowledge demand feature analysis model based on the feedback data.

Images