In the era of artificial intelligence,cognitive computing,based on cognitive science;and supported by machine learning and big data,brings personalization into every corner of our social life.Recommendation systems ar...In the era of artificial intelligence,cognitive computing,based on cognitive science;and supported by machine learning and big data,brings personalization into every corner of our social life.Recommendation systems are essential applications of cognitive computing in educational scenarios.They help learners personalize their learning better by computing student and exercise characteristics using data generated from relevant learning progress.The paper introduces a Learning and Forgetting Convolutional Knowledge Tracking Exercise Recommendation model(LFCKT-ER).First,the model computes students’ability to understand each knowledge concept,and the learning progress of each knowledge concept,and the model consider their forgetting behavior during learning progress.Then,students’learning stage preferences are combined with filtering the exercises that meet their learning progress and preferences.Then students’ability is used to evaluate whether their expectations of the difficulty of the exercises are reasonable.Then,the model filters the exercises that best match students’expectations again by students’expectations.Finally,we use a simulated annealing optimization algorithm to assemble a set of exercises with the highest diversity.From the experimental results,the LFCKT-ER model can better meet students’personalized learning needs and is more accurate than other exercise recommendation systems under various metrics on real online education public datasets.展开更多
The exercise recommendation system is emerging as a promising application in online learning scenarios,providing personalized recommendations to assist students with explicit learning directions.Existing solutions gen...The exercise recommendation system is emerging as a promising application in online learning scenarios,providing personalized recommendations to assist students with explicit learning directions.Existing solutions generally follow a collaborative filtering paradigm,while the implicit connections between students(exercises)have been largely ignored.In this study,we aim to propose an exercise recommendation paradigm that can reveal the latent connections between student-student(exercise-exercise).Specifically,a new framework was proposed,namely personalized exercise recommendation with student and exercise portraits(PERP).It consists of three sequential and interdependent modules:Collaborative student exercise graph(CSEG)construction,joint random walk,and recommendation list optimization.Technically,CSEG is created as a unified heterogeneous graph with students’response behaviors and student(exercise)relationships.Then,a joint random walk to take full advantage of the spectral properties of nearly uncoupled Markov chains is performed on CSEG,which allows for full exploration of both similar exercises that students have finished and connections between students(exercises)with similar portraits.Finally,we propose to optimize the recommendation list to obtain different exercise suggestions.After analyses of two public datasets,the results demonstrated that PERP can satisfy novelty,accuracy,and diversity.展开更多
Purpose-Recently,the number of online learners and learning resources has increased dramatically,and the knowledge network generated in the e-learning platform is getting vaster and more complex than ever.Analyzing le...Purpose-Recently,the number of online learners and learning resources has increased dramatically,and the knowledge network generated in the e-learning platform is getting vaster and more complex than ever.Analyzing learners’potential preferences by aggregating high-level semantic information from this network and accurately modeling their cognitive states is crucial for identifying similar learners.Combining similar learners’learning records helps recommend suitable exercises to improve the effectiveness of exercise recommendations.This article tackles the challenging problem of how to aggregate high-level semantic information in a huge graph and accurately model learners’cognitive states.Design/methodology/approach-Firstly,this approach constructs e-learning environments’knowledge graphs by integrating the difficulty of exercises and characteristics of answering behaviors,and the knowledge graph attention network(KGAT)is used to train the graph embedding model of the knowledge graph.Secondly,a score reevaluation method is designed based on the coefficient of completion quality to help accurately model learners’cognitive states.Then,the learners’actual cognitive states,obtained by the cognitive diagnosis model(CDM),are innovatively incorporated into graph matching for acquiring similar subgraphs.Finally,the personalized recommendation results are ranked according to learners’interaction probability on similar exercises.Findings-First,the proposed method has superior exercise recommendation performance.Experiments demonstrate that,compared to the existing approach,the proposed approach has an increase rate of 3.21%,3.32%,3.27%and 0.38%in precision,recall,F1 score and HR@10,respectively,in the large-scale graph data scenario.Second,aggregating high-level semantic information from the knowledge network helps explore learners’potential preferences.Finally,the fine-grained scoring mechanism based on learners’exercise completion quality can better reflect the actual mastery levels of learners,which improves the accuracy of modeling their cognitive states.Originality/value-First,an approach to personalized exercise recommendation is proposed via knowledge enhancement and fuzzy cognitive fusion.The experiments demonstrate the effectiveness and feasibility of this approach in a scenario with large-scale graph data.Second,this approach provides a flexible and adaptable framework.In it,the CDM can be replaced to explore for better accuracy of cognitive evaluation.Third,KGAT is employed to embed the knowledge graph in e-learning environments for aggregating high-level semantic information from the graph.Finally,a score reevaluation method is designed to analyze learners’learning behavior for accurately modeling their cognitive states.展开更多
基金supported by the National Natural Science Foundation of China(No.62006090)Research Funds of Central China Normal University(CCNU)under Grants 31101222211 and 31101222212.
文摘In the era of artificial intelligence,cognitive computing,based on cognitive science;and supported by machine learning and big data,brings personalization into every corner of our social life.Recommendation systems are essential applications of cognitive computing in educational scenarios.They help learners personalize their learning better by computing student and exercise characteristics using data generated from relevant learning progress.The paper introduces a Learning and Forgetting Convolutional Knowledge Tracking Exercise Recommendation model(LFCKT-ER).First,the model computes students’ability to understand each knowledge concept,and the learning progress of each knowledge concept,and the model consider their forgetting behavior during learning progress.Then,students’learning stage preferences are combined with filtering the exercises that meet their learning progress and preferences.Then students’ability is used to evaluate whether their expectations of the difficulty of the exercises are reasonable.Then,the model filters the exercises that best match students’expectations again by students’expectations.Finally,we use a simulated annealing optimization algorithm to assemble a set of exercises with the highest diversity.From the experimental results,the LFCKT-ER model can better meet students’personalized learning needs and is more accurate than other exercise recommendation systems under various metrics on real online education public datasets.
基金supported by the Industrial Support Project of Gansu Colleges under Grant No.2022CYZC-11Gansu Natural Science Foundation Project under Grant No.21JR7RA114+1 种基金National Natural Science Foundation of China under Grants No.622760736,No.1762078,and No.61363058Northwest Normal University Teachers Research Capacity Promotion Plan under Grant No.NWNU-LKQN2019-2.
文摘The exercise recommendation system is emerging as a promising application in online learning scenarios,providing personalized recommendations to assist students with explicit learning directions.Existing solutions generally follow a collaborative filtering paradigm,while the implicit connections between students(exercises)have been largely ignored.In this study,we aim to propose an exercise recommendation paradigm that can reveal the latent connections between student-student(exercise-exercise).Specifically,a new framework was proposed,namely personalized exercise recommendation with student and exercise portraits(PERP).It consists of three sequential and interdependent modules:Collaborative student exercise graph(CSEG)construction,joint random walk,and recommendation list optimization.Technically,CSEG is created as a unified heterogeneous graph with students’response behaviors and student(exercise)relationships.Then,a joint random walk to take full advantage of the spectral properties of nearly uncoupled Markov chains is performed on CSEG,which allows for full exploration of both similar exercises that students have finished and connections between students(exercises)with similar portraits.Finally,we propose to optimize the recommendation list to obtain different exercise suggestions.After analyses of two public datasets,the results demonstrated that PERP can satisfy novelty,accuracy,and diversity.
文摘Purpose-Recently,the number of online learners and learning resources has increased dramatically,and the knowledge network generated in the e-learning platform is getting vaster and more complex than ever.Analyzing learners’potential preferences by aggregating high-level semantic information from this network and accurately modeling their cognitive states is crucial for identifying similar learners.Combining similar learners’learning records helps recommend suitable exercises to improve the effectiveness of exercise recommendations.This article tackles the challenging problem of how to aggregate high-level semantic information in a huge graph and accurately model learners’cognitive states.Design/methodology/approach-Firstly,this approach constructs e-learning environments’knowledge graphs by integrating the difficulty of exercises and characteristics of answering behaviors,and the knowledge graph attention network(KGAT)is used to train the graph embedding model of the knowledge graph.Secondly,a score reevaluation method is designed based on the coefficient of completion quality to help accurately model learners’cognitive states.Then,the learners’actual cognitive states,obtained by the cognitive diagnosis model(CDM),are innovatively incorporated into graph matching for acquiring similar subgraphs.Finally,the personalized recommendation results are ranked according to learners’interaction probability on similar exercises.Findings-First,the proposed method has superior exercise recommendation performance.Experiments demonstrate that,compared to the existing approach,the proposed approach has an increase rate of 3.21%,3.32%,3.27%and 0.38%in precision,recall,F1 score and HR@10,respectively,in the large-scale graph data scenario.Second,aggregating high-level semantic information from the knowledge network helps explore learners’potential preferences.Finally,the fine-grained scoring mechanism based on learners’exercise completion quality can better reflect the actual mastery levels of learners,which improves the accuracy of modeling their cognitive states.Originality/value-First,an approach to personalized exercise recommendation is proposed via knowledge enhancement and fuzzy cognitive fusion.The experiments demonstrate the effectiveness and feasibility of this approach in a scenario with large-scale graph data.Second,this approach provides a flexible and adaptable framework.In it,the CDM can be replaced to explore for better accuracy of cognitive evaluation.Third,KGAT is employed to embed the knowledge graph in e-learning environments for aggregating high-level semantic information from the graph.Finally,a score reevaluation method is designed to analyze learners’learning behavior for accurately modeling their cognitive states.
