CN112149001B - Learning companion recommendation system and method - Google Patents
Learning companion recommendation system and method Download PDFInfo
- Publication number
- CN112149001B CN112149001B CN202010978188.2A CN202010978188A CN112149001B CN 112149001 B CN112149001 B CN 112149001B CN 202010978188 A CN202010978188 A CN 202010978188A CN 112149001 B CN112149001 B CN 112149001B
- Authority
- CN
- China
- Prior art keywords
- learning
- learner
- knowledge
- score
- weight
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 60
- 238000004364 calculation method Methods 0.000 claims abstract description 32
- 230000003993 interaction Effects 0.000 claims abstract description 29
- 230000008569 process Effects 0.000 claims abstract description 29
- 238000002372 labelling Methods 0.000 claims abstract description 16
- 230000000694 effects Effects 0.000 claims description 121
- 238000012360 testing method Methods 0.000 claims description 66
- 238000010586 diagram Methods 0.000 claims description 44
- 238000011156 evaluation Methods 0.000 claims description 36
- 238000013507 mapping Methods 0.000 claims description 25
- 238000012512 characterization method Methods 0.000 claims description 16
- 230000006399 behavior Effects 0.000 claims description 14
- 238000004458 analytical method Methods 0.000 claims description 9
- 230000001149 cognitive effect Effects 0.000 claims description 9
- 230000001186 cumulative effect Effects 0.000 claims description 7
- 238000013528 artificial neural network Methods 0.000 claims description 5
- 238000013480 data collection Methods 0.000 claims description 2
- 238000011160 research Methods 0.000 description 9
- 238000005516 engineering process Methods 0.000 description 8
- 239000013598 vector Substances 0.000 description 6
- 238000001914 filtration Methods 0.000 description 4
- 238000011161 development Methods 0.000 description 3
- 230000006870 function Effects 0.000 description 3
- 230000002452 interceptive effect Effects 0.000 description 2
- 239000000463 material Substances 0.000 description 2
- 238000012549 training Methods 0.000 description 2
- 101100136092 Drosophila melanogaster peng gene Proteins 0.000 description 1
- 230000006978 adaptation Effects 0.000 description 1
- 230000002776 aggregation Effects 0.000 description 1
- 238000004220 aggregation Methods 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 230000019771 cognition Effects 0.000 description 1
- 230000006998 cognitive state Effects 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 230000008451 emotion Effects 0.000 description 1
- 239000002360 explosive Substances 0.000 description 1
- 238000012905 input function Methods 0.000 description 1
- 230000007786 learning performance Effects 0.000 description 1
- 238000007726 management method Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000005065 mining Methods 0.000 description 1
- 230000006855 networking Effects 0.000 description 1
- 238000013441 quality evaluation Methods 0.000 description 1
- 230000000306 recurrent effect Effects 0.000 description 1
- 238000012827 research and development Methods 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 238000003860 storage Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/953—Querying, e.g. by the use of web search engines
- G06F16/9536—Search customisation based on social or collaborative filtering
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/10—Services
- G06Q50/20—Education
- G06Q50/205—Education administration or guidance
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Databases & Information Systems (AREA)
- Theoretical Computer Science (AREA)
- Tourism & Hospitality (AREA)
- Physics & Mathematics (AREA)
- Educational Administration (AREA)
- Educational Technology (AREA)
- General Physics & Mathematics (AREA)
- Strategic Management (AREA)
- General Business, Economics & Management (AREA)
- Primary Health Care (AREA)
- Marketing (AREA)
- Human Resources & Organizations (AREA)
- General Health & Medical Sciences (AREA)
- Economics (AREA)
- Health & Medical Sciences (AREA)
- Data Mining & Analysis (AREA)
- General Engineering & Computer Science (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Electrically Operated Instructional Devices (AREA)
Abstract
The invention provides a learning companion recommendation system and a learning companion recommendation method, wherein the system comprises an intelligent terminal and a server; the intelligent terminal is used for online learning interaction, and an online learning companion discovery system is operated in the server, and comprises an information labeling module, a learning interaction and data acquisition module, a knowledge state and knowledge structure representation calculation module and a similarity calculation and learning companion recommendation module. The invention is suitable for a general online learning platform, can dynamically represent the knowledge state and the knowledge structure of a learner in the online learning process, and can obtain the learning companion which is the most similar in the two dimensions of the knowledge state and the knowledge structure.
Description
Technical Field
The invention belongs to the field of online learning resource recommendation, in particular to a learning companion recommendation system and method.
Background
With the explosive growth of Web information, it has become increasingly difficult to quickly obtain truly useful and accurate information from the Web that meets the needs of a precise personality. One possible solution is to conduct personalized recommendation of resources, and through mining individual demands of users, further actively find appropriate resource services from massive information resources, and present the appropriate resource services to the users in an appropriate mode. The recommendation system typically matches the interest requirement information in the user model with the feature information in the recommendation object model, and at the same time performs computational screening by means of a corresponding recommendation algorithm, finds out the recommendation objects that the user may be interested in, and then presents to the user (Xu Hailing, wu Xiao, li Xiaodong, yan Baoping. Internet recommendation system compares research [ J ]. Software journal, 2009, 02:350-362.).
At present, personalized resource recommendation is widely applied to a plurality of fields, including electronic commerce websites, portal websites, professional service information, book information websites, social networking websites, online learning systems and the like. In the field of online learning systems, the personalized recommendation system mainly finds learning contents, presentation modes and learning paths meeting personalized requirements of learners through identifying and analyzing factors such as learning styles, interest preferences, emotion states, cognitive states and the like of the learners. The nature of personalized recommendations is an information filtering strategy aimed at predicting and pushing items that may be of interest to a user. With the continuous development of internet technology, pervasive computing and the like, researchers at home and abroad have conducted more and more intensive research on personalized recommendation technologies, and content-Based recommendations (Chen Haoyi. Web page personalized recommendation technology research [ D ] da: university of maritime, 2009.), collaborative filtering-Based recommendations (herlock, j.l., konstan, j.a., borchers, a., & Riedl, J. (1999, august) & An algorithmic framework for performing collaborative filtering.in Proceedings of the, nd annual international ACM SIGIR conference on Research and development in information retrieval (pp.230-237) & ACM are formed according to different implementation algorithms and modes; miao Jing research on personalized recommendation algorithm Based on extended neighbors and semantic tree [ D ]. Dalian: university, 2010, bobadilla, j.e.s.u.s., serradill a, f., & Hernando, a. (2009) Collaborative filtering adapted to recommender Systems of e-learning Systems,22 (4), 261-265, sarwar, b., karypis, g., konstan, J., & ril, J., & 2001, april), & Item-Based collaborative filtering recommendation algorithms, in Proceedings of the, th international conference on World Wide Web (pp.285-295), ACM., & associated rule-Based recommendations, knowledge-Based recommendations, utility-Based recommendations (Chen Haoyi, web page personalized recommendation technology research [ D ]. Da. ], university of major sea, 2009.),. Hybrid recommendations (Salehi, m., & mamadi, i.n. (2013) Hybrid recommendation approach for learning material Based on sequential pattern of the accessed material and the learner', kurz., [ 6, z.), [ 3, z. ], and science-48,57 ], 2007,7 (14) 3394-3398; li Yan research and application based on learning content personalized recommendation of curriculum ontology [ D ]. Wuhan: university of Huazhong, 2013; jiang Jiang, zhao Wei, du Xin, liang Ming. Personalized ontology learning resource recommendation study based on user model [ J ]. Chinese electrochemical education, 2010, (05): 106-111.) seven personalized recommendation core technologies.
In the online learning field, what needs to be recommended is learning resources (including digital resources and interpersonal resources) which can truly help and guide a learner to enter a next learning link, and the purpose of recommendation is to truly promote learning, so that more contents need to be considered in the recommendation design, including characteristics of the learner, characteristics of other learning peers in the system, specific learning situations and attribute characteristics of the learning resources, and learning resources are various in types, learning characteristics are continuously changed, and factors influencing learning are more, so that an algorithm and a strategy for personalized resource recommendation are complex.
From the implementation technology of the personalized recommendation, the essence is to implement the personalized recommendation mainly by using element matching between the digital resource and the user, wherein the user element is mainly expressed in learning preference and item scoring of the digital resource, the knowledge state and knowledge structure of the learner user are rarely focused at the same time, and the reference element is often a staged content element acquired on a certain time slice. Therefore, the study takes the knowledge state and knowledge structural elements of the learner as an important index for recommendation, and dynamically and continuously updates the recommended elements by combining the dynamic characteristics of the learning process.
The concept graph is widely applied and accepted in the field of education practice as a tool for evaluating learning performance of learners, and a plurality of researchers conduct intensive research on evaluation standards of the concept graph from two aspects of content and structure, so that a certain reference and reference are provided for representing the similarity of the cognitive map of the learners in the research. Novak and golin designed a set of criteria for manually evaluating learner conceptual diagrams: for each active connection score of 1, for each active ranking score of 5, for each active and important cross-connect score of 10, for each active and not embodying the cross-connect score of 2 for the comprehensive link between related concepts (propositions), for each example score of 1, the sum of these scores is the final score of the conceptual diagram (Novak, j.d., & golin, d.b. (1984). Liu and Lee designed a method for realizing automatic assessment of quality of conceptual diagrams by comparing the conceptual diagrams of learners with expert conceptual diagrams (Liu, s.h., & Lee, g.g., (2013). Using a concept map knowledge management system to enhance the learning of biology. Computers & products, 68, 105-116). It should be noted that the research of the quality evaluation criteria of the conceptual diagram brings great implications to the calculation based on knowledge states and knowledge structure similarity.
Disclosure of Invention
The invention aims to solve the technical problems that: based on the theory of the nearest neighbor development area of the Wiggust, the knowledge state and the knowledge structure of the learner are dynamically represented, learning peers which are the most similar in knowledge state and knowledge structure level are searched for by the learner, the solitary feeling in the learning process is reduced, and the learning peer recommendation system and method are provided, so that the flow is simple, the feedback is timely, the learning peers are dynamically updated, and the learning peer recommendation system and method are suitable for learning scenes of large-scale online courses.
According to one aspect of the invention, a learning companion recommendation system is provided, the online learning companion discovery system comprises an information labeling module, a learning interaction and data acquisition module, a knowledge state and knowledge structure characterization calculation module and a similarity calculation and learning companion recommendation module, wherein:
the information labeling module is used for establishing association relations and association weights between the subject knowledge points in the subject knowledge graph and learning activities, learning contents and exercise tests, and setting target levels required by the subject knowledge points in the learning activities, the learning contents and the exercise tests;
the learning interaction and data acquisition module is used for storing a behavior record generated by the learning interaction in the learning process interaction data record base, and the learning interaction comprises participation in learning activities, answer exercise test and operation learning content;
The knowledge state and knowledge structure calculation characterization module is used for calculating the score of the learner on the subject knowledge point, and comparing the score with the pass score set by the evaluation item in the evaluation scheme to obtain the knowledge state of the learner on the subject knowledge point; calculating the association weight between two discipline knowledge points according to the weight of the course unit in the whole online course, the weight of the learning activity in the course unit and the weight of the discipline knowledge points in the learning activity, so as to judge whether the association exists between the two discipline knowledge points, if so, continuously using the weight of the course unit in the online course, the score of the learner in the learning activity and the weight of the discipline knowledge points in the learning activity, and calculating the understanding degree weight of the learner on the association structure between the two discipline knowledge points to obtain the knowledge structure of the learner about the associated discipline knowledge points;
the similarity calculation and learning companion recommendation module is used for constructing a learning state diagram of the learner by taking the subject knowledge points related to the online courses as vertexes and the association relations existing between the subject knowledge points as edges, mapping the knowledge state and the knowledge structure of the learner into state information of the vertexes and the edges in the learning state diagram of the learner, calculating the similarity between the learning state diagrams, and obtaining the recommended learning companion with high similarity.
Furthermore, in the information labeling module, the target hierarchy is set according to six hierarchies specified in the education target classification theory of the revision bloom with respect to the cognitive domain, including memorization, understanding, application, analysis, evaluation and creation.
Further, in the learning interaction and data collection module, the operation learning content includes nine types of browsing learning content, editing learning content, annotating learning content, subscribing learning content, sharing learning content, collecting learning content, downloading learning content, uploading learning content and comment learning content.
Further, in the knowledge state and knowledge structure calculation characterization module, the method for calculating the score of the learner on the subject knowledge point is as follows:
(1) Setting the weight of the course unit in the online course and the weight of the learning activity in the course unit, setting the weight of three evaluation items of participation in the learning activity, answer exercise test and operation learning content in the course unit in an evaluation scheme, and requesting the learner to obtain a percentage and a grid score;
(2) The score of the learner on the subject knowledge points is calculated in a cumulative weighting mode by utilizing the association weights between the subject knowledge points and learning activities, exercise tests and learning contents and required target levels.
Further, the knowledge states are two states of mastered state and unclamped state, and the knowledge structure of the subject knowledge points is the association established and the association not established.
According to another aspect of the present invention, a learning companion recommendation method is provided, which is characterized by comprising the steps of:
establishing an association relation between the subject knowledge points in the subject knowledge graph and the learning activities, the learning contents and the exercise tests, setting association weights occupied by the subject knowledge points in the learning activities, the learning contents and the exercise tests, and setting target levels required by the subject knowledge points in the learning activities, the learning contents and the exercise tests;
step (2) students perform online course learning, wherein the online course learning comprises participation in learning activities, answer exercise testing and operation of learning content, and behavior records generated by learning are stored in a learning process interaction data record base;
setting the weight occupied by the course unit in the online course and the weight occupied by the learning activity in the course unit, setting the weights occupied by three evaluation items of participation in the learning activity, answer exercise test and operation learning content in the course unit in an evaluation scheme, and requesting a learner to obtain a percentage and a grid score;
Step (4) automatically scoring two evaluation items of answer exercise test and operation learning content of a learner according to the data record stored in the learning process interaction data record module; scoring the evaluation items of the learner in the learning activities by the learning companion and the online course teacher; calculating the final score of the learner on the subject knowledge point in a cumulative weighting mode;
step (5) utilizing the weight omega of course units in online courses ui Weights ω of learning activities in course units aui Weights ω of discipline knowledge points in learning activities kai Respectively calculating the association weight RC between two discipline knowledge points; when the association weight is larger than a preset association threshold delta, indicating that association exists between the subject knowledge points, continuously calculating an understanding degree weight URC of the learner on an association structure between the two subject knowledge points by using the weight of the course unit in the whole online course, the score of the learner in the learning activity and the weight of the subject knowledge points in the learning activity; comparing the understanding degree weight with a preset understanding threshold eta, and further obtaining a knowledge structure of the learner on two discipline knowledge points;
And (6) constructing a learning state diagram of the learner by taking the subject knowledge points related to the online courses as vertexes and the association relations existing between the subject knowledge points as edges, mapping the knowledge state and the knowledge structure of the learner into state information of the vertexes and the edges in the learning state diagram of the learner, and calculating the similarity between the learning state diagrams, wherein the similarity is high and is a recommended learning companion.
Further, in the step (1), the target hierarchy is set according to six hierarchies specified in the education target classification theory of the revision bloom with respect to the cognitive domain, including memorization, understanding, application, analysis, evaluation, and creation.
Further, the step (4) includes:
step 41, automatically scoring two evaluation items of answer exercise test and operation learning content of a learner according to data records stored in a learning process interaction data record base;
step 42, for scoring participation in learning activities, five students respectively score in a percentage system, namely Ea, eb and Ec after the highest score and the lowest score are removed, when abs (Ea, eb) is less than or equal to thr, abs (Ea, ec) is less than or equal to thr, abs (Eb, ec) is less than or equal to thr, thr is a threshold set by online course teachers, and the score of the learner participating in the learning activities is (Ea+Eb+ec)/3; otherwise, the score of the learner participating in the learning activity is given by an online course teacher and is recorded as E0; abs represents absolute value;
Step 43, calculating the final score f=0.01× (sai×ω kai ×ε kai +STi×ω kti ×ε kti +SCi×ω kci ×ε kci )/(ω kai ×ε kai +ω kti ×ε kti +ω kci ×ε kci ) And a lattice weighting value c=0.01× (paui×ω kai ×ε kai +Ptui×ω kti ×ε kti +Pcui×ω kci ×ε kci ) /(Paui+Ptui+Pcui); when F is more than or equal to C, mapping the knowledge state of the learner on the subject knowledge point to be mastered, otherwise mapping to be not mastered;
wherein SAi, STi, SCi is the score, ω of the learner in participating in learning activities, doing answer exercise test, manipulating learning content kai 、ω kti 、ω kci For the association weight between the subject knowledge points and learning activities, exercise tests and learning contents kai 、ε kti 、ε kci For the target level, the final score f=0.01× (sai×ω kai ×ε kai +STi×ω kti ×ε kti +SCi×ω kci ×ε kci )/(ω kai ×ε kai +ω kti ×ε kti +ω kci ×ε kci ) Paui, ptui, pcui are, in order, a percentile score for participating in a learning activity, a percentile score for answering an exercise test, and a percentile score for operating a learning content.
Further, the step (5) includes:
step 51, using the weight ω of course unit in the whole online course ui Weights ω of learning activities in course units aui Weights ω of discipline knowledge points in learning activities kai Calculating the association weight RC (KP) m ，KP n ) The method comprises the steps of carrying out a first treatment on the surface of the If omega kam > 0 and omega kan ＞0，RC(KP m ，KP n )＝∑ u ∑ a (ω ui ×(100×ω aui )×(ω kam +ω kan )/∑ a (100×ω aui ) Or RC (KP) m ，KP n ) =0, where u is the number of course units in the online course, and a is the number of learning activities in the course units; when RC (KP) m ，KP n ) If not, then continue to calculate the learner's understanding level URC (KP) m ，KP n )＝∑ u ∑ a ((ω ui ×(SA×ω aui )×(ω kam +ω kan )/∑ a (100×ω aui ))/(ω ui ×(100×ω aui )×(ω kam +ω kan )/∑ a (100×ω aui ) A) is set forth; when URC (KP) m ，KP n ) And when the relation is not less than eta, mapping the knowledge structures of the two discipline knowledge points by the learner into established relations, otherwise, mapping into unexplained relations.
Further, in the step (6), similarity between the learning state diagrams is calculated through the graph neural network.
The technical scheme of the invention is specifically described below: a learning companion recommendation system and method, the said system includes intelligent terminal, server; the online learning companion discovery system is operated in the server and comprises an information labeling module, a learning interaction and data acquisition module, a knowledge state and knowledge structure representation calculation module and a similarity calculation and learning companion recommendation module, wherein:
the information labeling module: according to an educational discipline knowledge graph constructed by discipline teachers or experts, establishing association relations between discipline knowledge points in the discipline knowledge graph, learning activities, learning contents and exercise tests by the online course teachers, and setting association weights of the discipline knowledge points in the learning activities, the learning contents and the exercise tests; meanwhile, according to six levels of memory, understanding, application, analysis, evaluation and creation specified in the educational objective classification theory of the revision bloom on the cognitive domain, setting objective levels required by subject knowledge points in learning activities, learning contents and exercise tests; the key attributes of the information labeling module comprise subject knowledge point names, associated resource information types, associated resource information IDs, associated weights and subject knowledge point target levels;
Learning interaction and data acquisition module: students use computers or intelligent mobile terminal equipment to perform online course learning, wherein the online course learning process comprises participation in learning activities, answer exercise testing and operation of learning content, and meanwhile, behavior records generated in the online course learning process are stored in an interaction data record base of the learning process; the operation learning content comprises nine types of browsing learning content, editing learning content, annotating learning content, subscribing learning content, sharing learning content, collecting learning content, downloading learning content, uploading learning content and comment learning content; all contents are recorded in the form of 'operation main body-operation type-operation content-operation start-stop time', and recorded data are stored in a behavior database so as to facilitate interactive behavior analysis;
knowledge state and knowledge structure calculation characterization module: the online course teacher sets the weight occupied by the course unit in the whole online course and the weight occupied by the learning activity in the course unit, sets the weights occupied by three evaluation items of the learning activity participation, the answer exercise test and the operation learning content in the evaluation scheme and the percentage and grid score required to be obtained by a learner, wherein the learner scores of the two evaluation items of the answer exercise test and the operation learning content are automatically scored according to the data records stored in the learning interaction and acquisition module, and the learner scores of the evaluation items participating in the learning activity are scored by a learning companion and the online course teacher together; in the process of scoring the participation learning activities by the learning companion and the course teacher together, each participation learning activity is firstly independently scored by five students in percentage, then one of the highest score and the lowest score is removed, and when the error between the remaining three scores is within the allowable threshold set by the online course teacher, the score of the participation learning activity of the learner is obtained by calculating the average value of the three scores; when the error between the three remaining scores exceeds an allowable threshold set by an online course teacher, the score of the learner participating in the learning activity is subjected to a percentile score by the online course teacher, and the score of the online course teacher is used as the score of the learner participating in the learning activity; calculating the final score of a learner on a specific subject knowledge point in a cumulative weighting mode by utilizing the association weight between the subject knowledge point and learning activity, exercise test and learning content stored in the information labeling module and a required target level, comparing the final score with pass score weighted values set in three evaluation items of participation learning activity, answer exercise test and operation learning content associated with the subject knowledge point by an online course teacher, and mapping the final score into the knowledge state of the learner on the subject knowledge point, wherein the knowledge state comprises mastered and not mastered two states; meanwhile, calculating the association weight between two different academic knowledge points by using the weight of a course unit in the whole online course, the weight of a learning activity in the course unit and the weight of a academic knowledge point in the learning activity set by an online course teacher; when the calculated association weight is larger than the association threshold value set by the course teacher, indicating that the association exists between the two discipline knowledge points, continuously calculating the understanding degree weight of the learner on the association structure between the two discipline knowledge points by using the weight of the course unit set by the online course teacher in the whole online course, the score of the learner in the learning activity and the weight of the discipline knowledge points in the learning activity; when the calculated understanding degree weight is compared with an understanding threshold value set by an online course teacher, mapping the understanding degree weight into a knowledge structure of the learner on the two discipline knowledge points, wherein the knowledge structure comprises established association and unexploited association;
Similarity calculation and learning companion recommendation module: and constructing a learning state diagram of the learner by taking the subject knowledge points related to the whole online course as vertexes and the association relations existing among the subject knowledge points as edges, mapping the knowledge state and the knowledge structure of the learner obtained by the knowledge state and knowledge structure calculation characterization module into state information of the vertexes and the edges of the learning state diagram, and calculating the similarity among different learning state diagrams by utilizing the ideas of the map neural network. According to the similarity value of the learning state diagram, the learning state diagram which is most matched with the current learning state diagram on the two layers of the knowledge state and the knowledge structure is found, and then the learning companion behind the learning state diagram is found, so that the learning companion is recommended to the current learner.
The intelligent terminal is a mobile phone or a tablet personal computer provided with an online learning companion discovery system client.
Compared with the prior art, the invention has the beneficial effects that:
(1) According to the invention, through establishing the association between the knowledge points of the education subjects in the knowledge graph and the learning activities, exercise tests and learning contents in the online courses, the interactive process data acquisition facing the learning behaviors and the cognitive input of the learner in the online learning scene is realized. The former technology is mainly oriented to the acquisition of learning behavior data in an online learning scene (Pafin, han Peng, wang Shaoqing, miao Xiaolong. Personalized recommendation method and system of learning behavior [ P ]. Chongqing: CN105404687A,2016-03-16; zhu Haiping, chen Yan, zheng Qinghua, bao Hongying, tian Feng. A learning behavior and efficiency analysis method oriented to a knowledge map [ P ]. Shanxi: CN104484454A, 2015-04-01.), but the consideration of the acquisition of cognition input level data is relatively poor.
(2) According to the theory of the nearest neighbor development area of the Wiggust, the similarity between knowledge states and knowledge structures of different learners obtained by dynamic characterization calculation is used as a basis for finding online learning peers, and the recommendation method and the process are designed from the perspectives of the learners and the learners, so that the suitability of the obtained learning peers is ensured, the method has the characteristics of dynamic adaptation and continuous updating, and the solitary feeling generated in the online course learning process can be reduced, which is not realized in the prior online learning recommendation technology.
Drawings
FIG. 1 is a flow chart of a learning companion recommendation system and method according to one embodiment of the present invention;
FIG. 2 is a flowchart of an implementation of an information labeling module according to an embodiment of the present invention;
FIG. 3 is a flow chart illustrating the implementation of knowledge state characterization computation in the knowledge structure characterization computation module, in accordance with an embodiment of the invention;
FIG. 4 is a flow chart of the knowledge structure characterization calculation in the knowledge state and knowledge structure characterization calculation module according to an embodiment of the invention;
fig. 5 is a flow chart illustrating a similarity calculation implementation according to an embodiment of the present invention.
Detailed Description
For the purpose of making the objects, technical solutions and advantages of the embodiments of the present invention more clear, the technical solutions of the embodiments of the present invention will be examined and fully described below with reference to the accompanying drawings in the embodiments of the present invention, and it is apparent that the described embodiments are some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
The invention will be described in detail with reference to the drawings and detailed description.
The invention provides a learning companion recommendation system and method, comprising an information labeling module, a learning interaction and data acquisition module, a knowledge state and knowledge structure representation calculation module and a similarity calculation and learning companion recommendation module, wherein the manufacturing flow is shown in figure 1. The specific steps are as follows:
as shown in fig. 2, the information labeling module in the present invention is specifically implemented as follows:
(1) According to the educational discipline knowledge graph constructed by discipline teacher or expert, on-line course teacher manually establishes the association relation between discipline knowledge points in discipline knowledge graph and learning activity, learning content and training test by using visual editing page, sets the association weight of different discipline knowledge points in learning activity, learning content and training test, wherein the weight omega of learning activity part discipline knowledge points ka1 +ω ka2 +…+ω kai =1, learning the weight ω of the content part discipline knowledge point kc1 +ω kc2 +…+ω kci =1, practice weights ω of test part discipline knowledge points kt1 +ω kt2 +…+ω kti =1; meanwhile, according to six levels of memory, understanding, application, analysis, evaluation and creation specified in the educational objective classification theory of the revision bloom on the cognitive domain, setting objective levels required by different academic knowledge points in learning activities, learning contents and exercise tests;
(2) Related data in the information labeling module are stored in an online course unit and learning activity, exercise test and learning content association table through an online course unit plurseUnit, a resource name, a resource type, a resource number resource ID, a resource association weight resource weight; meanwhile, through the subject knowledge roll name, the resource type, the resource number resource ID, the knowledge association weight knowledgeWeight, the target level targetLevel and the time timetable are stored in a subject knowledge point and learning activity, exercise test and learning content association table.
The learning interaction and data acquisition module in the invention is specifically realized as follows:
(1) Students use computers or intelligent mobile terminal equipment to perform online course learning, wherein the online course learning process comprises participation in learning activities, answer exercise testing and operation of learning content, and meanwhile, behavior records generated in the online course learning process are stored in an interaction data record base of the learning process; the operation learning content comprises nine types of browsing learning content, editing learning content, annotating learning content, subscribing learning content, sharing learning content, collecting learning content, downloading learning content, uploading learning content and comment learning content; the data storage fields of the process record table participating in the learning activity comprise learning activity IDs, learning activity participation start and stop time, learning activity task work storage path addresses and learner IDs; the stored data fields of the process record table for answering the exercise test comprise an exercise test ID, an exercise test answer start-stop time, exercise test answer content and a learner ID; the stored data fields of the process record table for operating the learning content include a learning content ID, a learning content operation start-stop time, a learning content operation type, and a learner ID.
As shown in fig. 3 and fig. 4, the knowledge state and knowledge structure characterization calculation module in the present invention is specifically implemented as follows:
(1) An online course teacher sets the weight omega occupied by course units in the whole online course ui Setting the weight omega occupied by the participation in the learning activities in the course unit aui And a percentile passing score P aui Setting the weight omega occupied by answer exercise test in course unit tui And a percentile passing score P tui Setting a weight omega occupied by the operation learning content in the course unit cui And a percentile passing score P cui Wherein omega u1 +ω u2 +…+ω ui ＝1，ω au1 +ω au2 +…+ω aui ＝1，ω tu1 +ω tu2 +…+ω tui ＝1，ω cu1 +ω cu2 +…+ω cui ＝1；
(2) The learner score on the answer practice test is automatically judged by the system according to the result of comparing the answer content of the students with the practice test standard answer provided by the online course teacher, and the score of all answer pairs is accumulated to obtain the total score ST of the learner on the practice test Ti; the scores on the operation learning content are automatically judged by a system according to the frequency ordering of nine operation behaviors of browsing learning content, editing learning content, annotating learning content, subscribing learning content, sharing learning content, collecting learning content, downloading learning content, uploading learning content and commenting learning content, and the scores SC= (1/browsing learning content frequency ordering+1/editing learning content frequency ordering+1/annotating learning content frequency ordering+1/subscribing learning content frequency ordering+1/sharing learning content frequency ordering+1/collecting learning content frequency ordering+1/downloading learning content frequency ordering+1/uploading learning content frequency ordering+1/commenting learning content frequency ordering) of 100/9; the scores of the study participation activities are scored by the study peers and online course teachers together, and each study participation activity is scored independently by five students in percentage system respectively, and is marked as E1, E2, E3, E4 and E5; then, max { E1, E2, E3, E4, E5} and Min { E1, E2, E3, E4, E5} are removed, ea, eb, ec are remained, when abs (Ea, eb) is less than or equal to thr and abs (Ea, ec) is less than or equal to thr and abs (Eb, ec) is less than or equal to thr, thr is a threshold set by an online course teacher, and the score SA of the learner participating in the learning activity is (Ea+Eb+ec)/3; otherwise, the score of the learner participating in the learning activity is scored by the online course teacher in a percentage way, and is marked as E0, and E0 is the score SA of the learner participating in the learning activity;
(3) The scores of the learners obtained in the above steps in the learning activities, the answer exercise tests and the operation learning content are marked as SAi, STi, SCi; and simultaneously, based on the association weights between the subject knowledge points stored in the information labeling module and learning activities, exercise tests and learning contents and required target levels, wherein the association weights are marked as omega kai 、ω kti 、ω kci The target level maps to ε kai 、ε kti 、ε kci And the mapping value is sequentially 1 to 6 from low to high according to six levels specified by the revision bloom educational objective; the final score f=0.01× (sai×ω) of the learner at a certain subject knowledge point is calculated by means of cumulative weighting kai ×ε kai +STi×ω kti ×ε kti +SCi×ω kci ×ε kci )/(ω kai ×ε kai +ω kti ×ε kti +ω kci ×ε kci ) Online course teachers participate in learning activities associated with the subject knowledge points,Answer practice test, and operational learning content, the pass score weighting value c=0.01× (paui×ω) kai ×ε kai +Ptui×ω kti ×ε kti +Pcui×ω kci ×ε kci ) /(Paui+Ptui+Pcui); when F is more than or equal to C, mapping the knowledge state of the learner on the subject knowledge point to be mastered, otherwise mapping to be not mastered;
(4) Weights ω of course units throughout online course set by online course teacher ui Weights ω of learning activities in course units aui Weights ω of discipline knowledge points in learning activities kai Calculating the association weight RC (KP) m ，KP n ) The method comprises the steps of carrying out a first treatment on the surface of the If omega kam > 0 and omega kan ＞0，RC(KP m ，KP n )＝∑ u ∑ a (ω ui ×(100×ω aui )×(ω kam +ω kan )/∑ a (100×ω aui ) Or RC (KP) m ，KP n ) =0, where u is the number of course units in the whole online course, and a is the number of learning activities in the course units; when RC (KP) m ，KP n ) When the learning level is more than or equal to delta, wherein delta is the association threshold value set by the online course teacher, the learning level URC (KP) of the association between the two discipline knowledge points of the learner is continuously calculated m ，KP n )＝∑ u ∑ a ((ω ui ×(SA×ω aui )×(ω kam +ω kan )/∑ a (100×ω aui ))/(ω ui ×(100×ω aui )×(ω kam +ω kan )/∑ a (100×ω aui ) A) is set forth; when URC (KP) m ，KP n ) And if not less than eta, wherein eta is an understanding threshold value set by an online course teacher, mapping the knowledge structures of the two discipline knowledge points into established association by a learner, and otherwise mapping into unexplained association.
As shown in fig. 5, the similarity calculation and learning companion recommendation module in the present invention is specifically implemented as follows:
taking the subject knowledge points related to the whole online course as vertexes and taking the between the subject knowledge pointsThe existing association relation is that a learning state diagram LSG= (V, E) of a learner is constructed for the edge and is expressed as a set of knowledge nodes V and edges E, and each node i epsilon V and a characteristic vector x i Associated with each edge (i, j) ∈E is a feature vector x ij Relatedness s between any two learning knowledge maps i，j ＝(LKG i ，LKG j ). The specific implementation process mainly comprises the following two links:
a) The knowledge nodes and edges in the learning state diagram are mapped into corresponding node vectors and edge vectors based on the multi-layer perceptron.
b) The node information is updated by using the propagation layer, and not only aggregation information on the edges of the learning state diagrams is considered, but also cross-diagram matching vectors for measuring the matching degree of one subject knowledge node in one learning state diagram and one or more subject knowledge nodes in other learning state diagrams are considered.
Wherein f message Is a connection input function on a typical multi-layer sensor, f match Is a function for cross-graph information calculation, f node Is a multi-layer perceptron or recurrent neural network, f LKG Is a layer representation calculation function, f s Is a standard spatial vector similarity calculation function.
Finally, according to the ordering of similarity values among different learning state diagrams, the learning state diagram which is most matched with the learning state diagram of the current learner on the two levels of the knowledge state and the knowledge structure is found, and then the learning companion behind the learning state diagram is found, so that the learning companion is recommended to the current learner.
According to another aspect of the invention, a learning companion recommendation method is provided, which comprises the following implementation steps:
according to an educational discipline knowledge graph constructed by discipline teachers or experts, an online course teacher establishes association relations between discipline knowledge points in the discipline knowledge graph and learning activities, learning contents and exercise tests, and association weights of the discipline knowledge points in the learning activities, the learning contents and the exercise tests are set; meanwhile, according to six levels of memory, understanding, application, analysis, evaluation and creation specified in the educational objective classification theory of the revision bloom on the cognitive domain, setting objective levels required by subject knowledge points in learning activities, learning contents and exercise tests;
Step (2) students use computers or intelligent mobile terminal equipment to perform online course learning, wherein the online course learning process comprises participation in learning activities, answer exercise tests and operation of learning content, and meanwhile, behavior records generated in the online course learning process are stored in an interaction data record base of the online course learning process; the operation learning content comprises nine types of browsing learning content, editing learning content, annotating learning content, subscribing learning content, sharing learning content, collecting learning content, downloading learning content, uploading learning content and comment learning content;
and (3) setting weights occupied by the course units in the whole online course, weights occupied by the learning activities in the course units by online course teachers, weights occupied by three evaluation items of participation in the learning activities, answer exercise test, operation learning content in the course units in an evaluation scheme and percentages and grid scores required to be obtained by learners.
Step (4) performing answer exercise test, operating learner scores of two types of evaluation items of learning content, automatically scoring according to a percentage system according to learning interaction and data records stored in a data acquisition module, and jointly scoring the learner scores of the evaluation items participating in learning activities by a learning companion and an online course teacher; in the scoring process of the learning activities of the learning companion and the course teacher, each learning activity is scored independently by five students in percentage system, and is marked as E1, E2, E3, E4 and E5; then, max { E1, E2, E3, E4, E5} and Min { E1, E2, E3, E4, E5} are removed, ea, eb, ec are remained, when abs (Ea, eb) is less than or equal to thr and abs (Ea, ec) is less than or equal to thr and abs (Eb, ec) is less than or equal to thr, thr is a threshold set by an online course teacher, and the score of the learner participating in the learning activity is (Ea+Eb+ec)/3; otherwise, the score of the learner participating in the learning activity is scored by the online course teacher in a percentage way, and is marked as E0, and E0 is the score of the learner participating in the learning activity;
Using the score of the learner obtained in the step (3) in participating in learning activities, answering exercise tests and operating learning contents, and marking as SAi, STi, SCi; based on the association weight between the subject knowledge points and the learning activities, the exercise test and the learning content and the required target level, wherein the association weight is marked as omega kai 、ω kti 、ω kci The target level maps to ε kai 、ε kti 、ε kci And the mapping value is sequentially 1 to 6 from low to high according to six levels specified by the revision bloom educational objective; computing learners by means of cumulative weightingThe final score f=0.01× (sai×ω kai ×ε kai +STi×ω kti ×ε kti +SCi×ω kci ×ε kci )/(ω kai ×ε kai +ω kti ×ε kti +ω kci ×ε kci ) The pass score weighting value c=0.01× (Paui×ω) set by the online course teacher in three kinds of evaluation items of participating in learning activities, answering exercise tests, and operating learning contents associated with the subject knowledge point kai ×ε kai +Ptui×ω kti ×ε kti +Pcui×ω kci ×ε kci ) /(Paui+Ptui+Pcui); when F is more than or equal to C, mapping the knowledge state of the learner on the subject knowledge point to be mastered, otherwise mapping to be not mastered; the method is consistent with the method in the knowledge state and knowledge structure characterization calculation module, and is not repeated.
Step (5) weighting omega of course units in the whole online course by using the online course teacher ui Weights ω of learning activities in course units aui Weights ω of discipline knowledge points in learning activities kai Respectively calculating the association weights RC between two different scientific knowledge points; when the calculated association weight is larger than an association threshold delta set by a course teacher, indicating that association exists between the two discipline knowledge points, continuously calculating an understanding degree weight URC of the learner on an association structure between the two discipline knowledge points by using the weight of a course unit set by the online course teacher in the whole online course, the score of the learner in the learning activity and the weight of the discipline knowledge points in the learning activity; when the calculated understanding degree weight is compared with an understanding threshold eta set by an online course teacher, mapping the understanding degree weight into a knowledge structure of the two subject knowledge points of a learner, wherein the knowledge structure comprises two situations of established knowledge association and unestablished knowledge association; the method is consistent with the method in the knowledge state and knowledge structure characterization calculation module, and is not repeated.
And (6) constructing a learning state diagram of the learner by taking subject knowledge points related to the whole online course as vertexes and taking association relations existing among the subject knowledge points as edges, and simultaneously mapping the knowledge state of the learner obtained in the step (4) and the knowledge structure of the learner obtained in the step (5) into state information of the vertexes and the edges of the learning state diagram respectively, and calculating the similarity among different learning state diagrams by using the ideas of a graph neural network. According to the similarity value of the learning state diagram, the learning state diagram which is most matched with the current learning state diagram on the two layers of the knowledge state and the knowledge structure is found, and then the learning companion behind the learning state diagram is found, so that the learning companion is recommended to the current learner. The method is consistent with the method in the similarity calculation and learning partner recommendation module, and is not repeated.
The most similar learning companion in the step (6) can continuously and dynamically change the current knowledge state and knowledge structure of the learner, and has the characteristic of real-time performance.
Parts of the invention not described in detail are well known in the art.
While the foregoing describes illustrative embodiments of the present invention to facilitate an understanding of the present invention by those skilled in the art, it should be understood that the present invention is not limited to the scope of the embodiments, but is to be construed as protected by the accompanying claims insofar as various changes are within the spirit and scope of the present invention as defined and defined by the appended claims.
Claims (10)
1. A learning companion recommendation system, characterized by: the system comprises an information labeling module, a learning interaction and data acquisition module, a knowledge state and knowledge structure characterization calculation module and a similarity calculation and learning companion recommendation module, wherein:
the information labeling module is used for establishing association relations and association weights between the subject knowledge points in the subject knowledge graph and learning activities, learning contents and exercise tests, and setting target levels required by the subject knowledge points in the learning activities, the learning contents and the exercise tests;
The learning interaction and data acquisition module is used for storing behavior records generated by the learning interaction in a learning process interaction data record base, wherein the learning interaction comprises participation in learning activities, answer exercise testing and operation learning content;
the knowledge state and knowledge structure calculation characterization module is used for calculating the score of the learner on the subject knowledge point, and comparing the score with the pass score set by the evaluation item in the evaluation scheme to obtain the knowledge state of the learner on the subject knowledge point; calculating the association weight between two discipline knowledge points according to the weight of the course unit in the whole online course, the weight of the learning activity in the course unit and the weight of the discipline knowledge points in the learning activity, so as to judge whether the association exists between the two discipline knowledge points, if so, continuously using the weight of the course unit in the online course, the score of the learner in the learning activity and the weight of the discipline knowledge points in the learning activity, and calculating the understanding degree weight of the learner on the association structure between the two discipline knowledge points to obtain the knowledge structure of the learner about the associated discipline knowledge points;
the similarity calculation and learning companion recommendation module is used for constructing a learning state diagram of the learner by taking the subject knowledge points related to the online courses as vertexes and the association relations existing between the subject knowledge points as edges, mapping the knowledge state and the knowledge structure of the learner into state information of the vertexes and the edges in the learning state diagram of the learner, calculating the similarity between the learning state diagrams, and obtaining the recommended learning companion with high similarity.
2. The system of claim 1, wherein the information labeling module wherein the target hierarchy is configured according to six hierarchies defined in the educational objective classification theory of the revision bloom with respect to the cognitive domain, including memory, understanding, application, analysis, evaluation, and creation.
3. The system of claim 1, wherein the learning interaction and data collection module is configured to operate learning content in nine types of learning content, including browsing learning content, editing learning content, annotating learning content, subscribing learning content, sharing learning content, collecting learning content, downloading learning content, uploading learning content, and commenting learning content.
4. The system of claim 1, wherein in the knowledge state and knowledge structure calculation characterization module, the method of calculating the learner's score at the subject knowledge point is:
(1) Setting the weight of the course unit in the online course and the weight of the learning activity in the course unit, setting the weight of three evaluation items of participation in the learning activity, answer exercise test and operation learning content in the course unit in an evaluation scheme, and requesting the learner to obtain a percentage and a grid score;
(2) The score of the learner on the subject knowledge points is calculated in a cumulative weighting mode by utilizing the association weights between the subject knowledge points and learning activities, exercise tests and learning contents and required target levels.
5. The system of claim 1, wherein the knowledge states are two states, mastered and not mastered, and the knowledge structure of the subject knowledge points is an established association and an unexploited association.
6. The learning companion recommendation method is characterized by comprising the following steps:
establishing an association relation between the subject knowledge points in the subject knowledge graph and the learning activities, the learning contents and the exercise tests, setting association weights occupied by the subject knowledge points in the learning activities, the learning contents and the exercise tests, and setting target levels required by the subject knowledge points in the learning activities, the learning contents and the exercise tests;
step (2) students perform online course learning, wherein the online course learning comprises participation in learning activities, answer exercise testing and operation of learning content, and behavior records generated by learning are stored in a learning process interaction data record base;
setting the weight occupied by the course unit in the online course and the weight occupied by the learning activity in the course unit, setting the weights occupied by three evaluation items of participation in the learning activity, answer exercise test and operation learning content in the course unit in an evaluation scheme, and requesting a learner to obtain a percentage and a grid score;
Step (4) automatically scoring two evaluation items of answer exercise test and operation learning content of a learner according to the data record stored in the learning process interaction data record module; scoring the evaluation items of the learner in the learning activities by the learning companion and the online course teacher; calculating the final score of the learner on the subject knowledge point in a cumulative weighting mode;
step (5) utilizing the weight omega of course units in online courses ui Weights ω of learning activities in course units aui Weights ω of discipline knowledge points in learning activities kai Respectively calculating the association weight RC between two discipline knowledge points; when the association weight is larger than a preset association threshold delta, indicating that association exists between the subject knowledge points, continuously calculating an understanding degree weight URC of the learner on an association structure between the two subject knowledge points by using the weight of the course unit in the whole online course, the score of the learner in the learning activity and the weight of the subject knowledge points in the learning activity; comparing the understanding degree weight with a preset understanding threshold eta, and further obtaining a knowledge structure of the learner on two discipline knowledge points;
And (6) constructing a learning state diagram of the learner by taking the subject knowledge points related to the online courses as vertexes and the association relations existing between the subject knowledge points as edges, mapping the knowledge state and the knowledge structure of the learner into state information of the vertexes and the edges in the learning state diagram of the learner, and calculating the similarity between the learning state diagrams, wherein the similarity is high and is a recommended learning companion.
7. The method of claim 6, wherein in step (1), the target hierarchy is set according to six hierarchies specified in the educational objective classification theory of the revision bloom with respect to the cognitive domain, including memorization, understanding, application, analysis, evaluation, and creation.
8. The method of claim 6, wherein the step (4) includes:
step 41, automatically scoring two evaluation items of answer exercise test and operation learning content of a learner according to data records stored in a learning process interaction data record base;
step 42, for scoring participation in learning activities, five students respectively score in a percentage system, namely Ea, eb and Ec after the highest score and the lowest score are removed, when abs (Ea, eb) is less than or equal to thr, abs (Ea, ec) is less than or equal to thr, abs (Eb, ec) is less than or equal to thr, thr is a threshold set by online course teachers, and the score of the learner participating in the learning activities is (Ea+Eb+ec)/3; otherwise, the score of the learner participating in the learning activity is given by an online course teacher and is recorded as E0; abs represents absolute value;
Step 43, calculating the final score f=0.01× (sai×ω kai ×ε kai +STi×ω kti ×ε kti +SCi×ω kci ×ε kci )/(ω kai ×ε kai +ω kti ×ε kti +ω kci ×ε kci ) And a lattice weighting value c=0.01× (paui×ω kai ×ε kai +Ptui×ω kti ×ε kti +Pcui×ω kci ×ε kci ) /(Paui+Ptui+Pcui); when F is more than or equal to C, mapping the knowledge state of the learner on the subject knowledge point to be mastered, otherwise mapping to be not mastered;
wherein SAi, STi, SCi is the score, ω of the learner in participating in learning activities, doing answer exercise test, manipulating learning content kai 、ω kti 、ω kci For the association weight between the subject knowledge points and learning activities, exercise tests and learning contents kai 、ε kti 、ε kci For the target level, the final score f=0.01× (sai×ω kai ×ε kai +STi×ω kti ×ε kti +SCi×ω kci ×ε kci )/(ω kai ×ε kai +ω kti ×ε kti +ω kci ×ε kci ) Paui, ptui, pcui are, in order, a percentile score for participating in a learning activity, a percentile score for answering an exercise test, and a percentile score for operating a learning content.
9. The method of claim 6, wherein the step (5) comprises:
step 51, using the weight ω of course unit in the whole online course ui Weights ω of learning activities in course units aui Weights ω of discipline knowledge points in learning activities kai Calculating the association weight RC (KP) m ，KP n ) The method comprises the steps of carrying out a first treatment on the surface of the If omega kam > 0 and omega kan ＞0，RC(KP m ，KP n )＝∑ u ∑ a (ω ui ×(100×ω aui )×(ω kam +ω kan )/∑ a (100×ω aui ) Or RC (KP) m ，KP n ) =0, where u is the number of course units, a is the number of learning activities in the course units; when RC (KP) m ，KP n ) When not less than delta, calculating the understanding degree URC (KP) of the learner on the correlation between two subject knowledge points m ，KP n )＝∑u∑a((ω ui ×(SA×ω aui )×(ω kam +ω kan )/∑ a (100×ω aui ))/(ω ui ×(100×ω aui )×(ω kam +ω kan )/∑ a (100×ω aui ) A) is set forth; when URC (KP) m ，KP n ) And when the relation is not less than eta, mapping the knowledge structures of the two discipline knowledge points by the learner into established relations, otherwise, mapping into unexplained relations.
10. The method according to claim 6, wherein in the step (6), the similarity between the learning state diagrams is calculated through a graph neural network.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010978188.2A CN112149001B (en) | 2020-09-17 | 2020-09-17 | Learning companion recommendation system and method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010978188.2A CN112149001B (en) | 2020-09-17 | 2020-09-17 | Learning companion recommendation system and method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112149001A CN112149001A (en) | 2020-12-29 |
| CN112149001B true CN112149001B (en) | 2023-04-25 |
Family
ID=73892370
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010978188.2A Active CN112149001B (en) | 2020-09-17 | 2020-09-17 | Learning companion recommendation system and method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112149001B (en) |
Families Citing this family (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112733046A (en) * | 2021-01-22 | 2021-04-30 | 西北民族大学 | Intelligent learning partner recommendation system based on college student electronic learning archives |
| CN113051300A (en) * | 2021-03-05 | 2021-06-29 | 深圳市鹰硕技术有限公司 | Online learning method and device based on learning partner matching |
| CN112883723A (en) * | 2021-03-18 | 2021-06-01 | 江西师范大学 | Deep neural network cognition level evaluation model based on Broumm cognition classification theory |
| CN113204717A (en) * | 2021-05-27 | 2021-08-03 | 广州大学 | Online learning community learning companion recommendation method and system based on context awareness |
| CN113742591B (en) * | 2021-09-07 | 2023-07-25 | 平安科技（深圳）有限公司 | Learning partner recommendation method and device, electronic equipment and storage medium |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106205248A (en) * | 2016-08-31 | 2016-12-07 | 北京师范大学 | A kind of representative learning person generates system and method at the on-line study cognitive map of domain-specific knowledge learning and mastering state |
| CN106503172A (en) * | 2016-10-25 | 2017-03-15 | 天闻数媒科技（湖南）有限公司 | The method and apparatus that learning path recommended by knowledge based collection of illustrative plates |
| CN107273490A (en) * | 2017-06-14 | 2017-10-20 | 北京工业大学 | A kind of combination mistake topic recommendation method of knowledge based collection of illustrative plates |
| CN108052672A (en) * | 2017-12-29 | 2018-05-18 | 北京师范大学 | Promote structural knowledge map construction system and method using group study behavior |
| WO2018155931A1 (en) * | 2017-02-24 | 2018-08-30 | 스케이프플로우 | Knowledge map-based multi-user game service providing device and method |
| CN108628967A (en) * | 2018-04-23 | 2018-10-09 | 西安交通大学 | A kind of e-learning group partition method generating network similarity based on study |
| CN109829059A (en) * | 2019-01-18 | 2019-05-31 | 平安科技（深圳）有限公司 | Recommend method, apparatus, equipment and the storage medium of knowledge point |
-
2020
- 2020-09-17 CN CN202010978188.2A patent/CN112149001B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106205248A (en) * | 2016-08-31 | 2016-12-07 | 北京师范大学 | A kind of representative learning person generates system and method at the on-line study cognitive map of domain-specific knowledge learning and mastering state |
| CN106503172A (en) * | 2016-10-25 | 2017-03-15 | 天闻数媒科技（湖南）有限公司 | The method and apparatus that learning path recommended by knowledge based collection of illustrative plates |
| WO2018155931A1 (en) * | 2017-02-24 | 2018-08-30 | 스케이프플로우 | Knowledge map-based multi-user game service providing device and method |
| CN107273490A (en) * | 2017-06-14 | 2017-10-20 | 北京工业大学 | A kind of combination mistake topic recommendation method of knowledge based collection of illustrative plates |
| CN108052672A (en) * | 2017-12-29 | 2018-05-18 | 北京师范大学 | Promote structural knowledge map construction system and method using group study behavior |
| CN108628967A (en) * | 2018-04-23 | 2018-10-09 | 西安交通大学 | A kind of e-learning group partition method generating network similarity based on study |
| CN109829059A (en) * | 2019-01-18 | 2019-05-31 | 平安科技（深圳）有限公司 | Recommend method, apparatus, equipment and the storage medium of knowledge point |
Non-Patent Citations (8)
| Title |
|---|
| Hai Zhuge 等.Knowledge Map: Mathematical Model and Dynamic Behaviors.《Journal of Computer Science and Technology》.2005,289–295. * |
| M Zhao 等.Effects of social and temporal distance on consumers' responses to peer recommendations.《Journal of Marketing Research》.1-41. * |
| 丁继红.个性化教育资源主动服务模式研究.《中国博士学位论文全文数据库社会科学Ⅱ辑》.2018,H127-1. * |
| 万海鹏 等.基于学习认知图谱的适应性学习框架构建与应用.《现代远距离教育》.2022,73-82. * |
| 刘昱良 等.基于科学知识图谱的学习路径推荐研究.《河南科技学院学报(自然科学版)》.2018,37-41. * |
| 刘昱良.基于知识图谱的个性化学习资源推荐研究.《天津市教科院学报 》.2019,I138-5494. * |
| 毕巧春.成人学习者的在线学习行为建模研究.《中国优秀硕士学位论文全文数据库社会科学Ⅱ辑》.2018,H127-103. * |
| 蔡霞 等.基于混合式学习的同伴推荐算法设计与实现.《计算机时代》.2020,38-40+44. * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112149001A (en) | 2020-12-29 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112149001B (en) | Learning companion recommendation system and method | |
| Albatayneh et al. | Utilizing learners’ negative ratings in semantic content-based recommender system for e-learning forum | |
| Yuan et al. | Homophily of network ties and bonding and bridging social capital in computer-mediated distributed teams | |
| Ghauth et al. | Measuring learner's performance in e-learning recommender systems | |
| Lai et al. | Novel personal and group-based trust models in collaborative filtering for document recommendation | |
| Li | Quality, evaluation and recommendation for learning object | |
| Haythornthwaite | Learning, connectivity and networks | |
| CN112131472A (en) | Information recommendation method and device, electronic equipment and storage medium | |
| Ranganath et al. | Facilitating time critical information seeking in social media | |
| Pipergias Analytis et al. | The structure of social influence in recommender networks | |
| Tang et al. | A literature review of personalized learning algorithm | |
| Wang | Exploration of data mining algorithms of an online learning behaviour log based on cloud computing | |
| CN116244513A (en) | Random group POI recommendation method, system, equipment and storage medium | |
| Sun et al. | Big educational data analytics, prediction and recommendation: a survey | |
| Wu et al. | Evaluating the E-Learning Platform from the Perspective of Knowledge Management: The AHP Approach. | |
| Buyrukoğlu | A novel color labeled student modeling approach using e-learning activities for data mining | |
| Jiang et al. | Augmenting personalized question recommendation with hierarchical information for online test platform | |
| Fan et al. | Personalized recommendation algorithm for curriculum-and politics-oriented hybrid teaching resources | |
| Paxinou et al. | Centrality Metrics from Students' Discussion Fora at Distance Education | |
| Shokeen | On measuring the role of social networks in project recommendation | |
| Shen et al. | [Retracted] International Chinese Education Expert System Based on Artificial Intelligence and Machine Learning Algorithms | |
| Li et al. | Design of Intelligent Push System for Teaching Resources Based on User Profile | |
| CN115080731A (en) | Comment language segment generation method and device, electronic equipment and storage medium | |
| Su et al. | A Cluster-based Personalized Item Recommended Approach on the Educational Assessment System. | |
| Burke et al. | Recommendation in the social web |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |