Knowledge Concept Recommendation Model for MOOCs with Local Subgraph Embedding

doi:10.3778/j.issn.1673-9418.2209056

Abstract

Abstract: Massive online open courses (MOOCs) have been extensively researched in reducing user learning blindness and improving user experience, especially personalized course resource recommendation based on graph neural networks. However, these efforts focus primarily on fixed or homogeneous graphs, vulnerable to data sparsity problems, and difficult to scale. This paper uses graph convolution on local subgraphs combined with an extended matrix factorization (MF) model to overcome this limitation. Firstly, the proposed method decomposes the heterogeneous graph into multiple meta-path-based subgraphs and combines random wandering sampling methods to capture complex semantic relationships between entities while sampling nodes’ influential neighborhoods, and performs graph convolution on local neighborhoods to smooth the representation of each node and achieve high scalability. Next, the attention mechanism adaptively fuses the contextual information of different subgraphs for a more comprehensive construction of user preferences. Finally, the model parameters are optimized by expanding MF to obtain recommendation list. To validate the performance of the proposed model, comparative experiments are conducted on publicly available MOOCs datasets, with a 2% performance improvement and a nearly 500% reduction in memory computation requirements compared with the optimal baseline, providing strong scalability while alleviating the data sparsity problem.

Key words: massive online open courses (MOOCs), graph neural networks, personalized course recommendation, graph convolution, meta-path-based subgraphs, extended matrix factorization

摘要： 大规模开放在线课程（MOOCs）在减少用户学习盲区和改善用户体验方面已经有大量的研究，尤其是基于图神经网络的个性化课程资源推荐，但现有工作主要集中在固定或同质图上，容易受到数据稀疏问题的影响且难以扩展。在局部子图上使用图卷积，并结合扩展的矩阵分解（MF）模型来解决这一问题。首先，将异构图分解为多个基于元路径的子图，结合随机游走采样方法实现在采样节点富有影响力邻域的同时捕获实体之间复杂的语义关系，并在局部邻域上进行图卷积平滑各节点表示，实现高可扩展性；然后，使用注意力机制适应性地融合不同子图的上下文信息，更全面地构建用户偏好；最后，通过扩展矩阵分解优化模型参数，获得推荐列表。为了验证提出模型的性能，在公开的MOOCs数据集上进行对比实验，相较于最优基线，性能提升了2%，内存计算需求降低了近500%，缓解数据稀疏问题的同时仍具有较强的可扩展性。

关键词: 大规模开放在线课程（MOOCs）, 图神经网络, 个性化课程推荐, 图卷积, 基于元路径的子图, 扩展矩阵分解

JU Chengcheng, ZHU Yi. Knowledge Concept Recommendation Model for MOOCs with Local Subgraph Embedding[J]. Journal of Frontiers of Computer Science and Technology, 2024, 18(1): 189-204.

居程程, 祝义. 采用局部子图嵌入的MOOCs知识概念推荐模型[J]. 计算机科学与探索, 2024, 18(1): 189-204.

References

[1] HUANG N, ZHANG J, BURTCH G, et al. Combating pro-crastination on massive online open courses via optimal calls to action[J]. Information Systems Research, 2021, 32(2): 301-317.
[2] ZHU M, SARI A, LEE M M. A systematic review of research methods and topics of the empirical MOOC literature (2014—2016)[J]. The Internet and Higher Education, 2018, 37: 31-39.
[3] SEATON D T, BERGNER Y, CHUANG I, et al. Who does what in a massive open online course?[J]. Communications of the ACM, 2014, 57(4): 58-65.
[4] KIZILCEC R F, PIECH C, SCHNEIDER E. Deconstructing disengagement: analyzing learner subpopulations in massive open online courses[C]//Proceedings of the 3rd International Conference on Learning Analytics and Knowledge, Leuven, Apr 8-12, 2013. New York: ACM, 2013: 170-179.
[5] JACOBSEN D Y. Dropping out or dropping in? A connectivist approach to understanding participants?? strategies in an e-learning MOOC pilot[J]. Technology, Knowledge and Learning, 2019, 24(1): 1-21.
[6] ADAMOPOULOS P. What makes a great MOOC? An inter-disciplinary analysis of student retention in online courses[C]//Proceedings of the 34th International Conference on Information Systems, Milano, Dec 15-18, 2013.
[7] KLOFT M, STIEHLER F, ZHENG Z, et al. Predicting MOOC dropout over weeks using machine learning methods[C]// Proceedings of the EMNLP 2014 Workshop on Analysis of Large Scale Social Interaction in MOOCs, Doha, Oct 25-29, 2014. Stroudsburg: ACL, 2014: 60-65.
[8] QIU J, TANG J, LIU T X, et al. Modeling and predicting learning behavior in MOOCs[C]//Proceedings of the 9th ACM International Conference on Web Search and Data Mining, San Francisco, Feb 22-25, 2016. New York: ACM, 2016: 93-102.
[9] NAGRECHA S, DILLON J Z, CHAWLA N V. MOOC dropout prediction: lessons learned from making pipelines interpretable[C]//Proceedings of the 26th International Conference on World Wide Web Companion, Perth, Apr 3-7, 2017: 351-359.
[10] CHEN Y, ZHANG M. MOOC student dropout: pattern and prevention[C]//Proceedings of the ACM Turing 50th Cele-bration Conference-China, Shanghai, May 12-14, 2017: 1-6.
[11] PANG Y, JIN Y, ZHANG Y, et al. Collaborative filtering re-commendation for MOOC application[J]. Computer Appli-cations in Engineering Education, 2017, 25(1): 120-128.
[12] BOUSBAHI F, CHORFI H. MOOC-Rec: a case based recom-mender system for MOOCs[J]. Procedia - Social and Behavioral Sciences, 2015, 195: 1813-1822.
[13] CAMPOS R, DOS SANTOS R P, OLIVEIRA J. A recom-mendation system based on knowledge gap identification in MOOCs ecosystems[C]//Proceedings of the XVI Brazilian Symposium on Information Systems, S?o Bernardo do Campo, Nov 3-6, 2020. New York: ACM, 2020: 1-8.
[14] LIU H, LI X. Learning path combination recommendation based on the learning networks[J]. Soft Computing, 2019, 24(6): 4427-4439.
[15] ZHANG M, ZHU J, WANG Z, et al. Providing personalized learning guidance in MOOCs by multi-source data analysis[J]. World Wide Web, 2018, 22(3): 1189-1219.
[16] LI J, YE Z. Course recommendations in online education based on collaborative filtering recommendation algorithm[J]. Complexity, 2020: 6619249.
[17] GONG J, WANG S, WANG J, et al. Attentional graph con-volutional networks for knowledge concept recommendation in MOOCs in a heterogeneous view[C]//Proceedings of the 43rd International ACM SIGIR Conference on Research and Development in Information Retrieval, Jul 25-30, 2020. New York: ACM, 2020: 79-88.
[18] GONG J, WANG C, ZHAO Z, et al. Automatic generation of meta-path graph for concept recommendation in MOOCs[J]. Electronics, 2021, 10(14): 1671.
[19] PIAO G. Recommending knowledge concepts on MOOC platforms with meta-path-based representation learning[C]// Proceedings of the 14th International Conference on Edu-cational Data Mining, Jun 29-Jul 2, 2021.
[20] GARG V, TIWARI R. Hybrid massive open online course (MOOC) recommendation system using machine learning[C]//Proceedings of the 2016 International Conference on Recent Trends in Engineering, Science & Technology, Hydera-bad, Oct 25-27, 2016: 1-5.
[21] FU D, LIU Q, ZHANG S, et al. The undergraduate-oriented framework of MOOCs recommender system[C]//Proceedings of the 2015 International Symposium on Educational Technology, Wuhan, Jul 27-29, 2015. Piscataway: IEEE, 2015: 115-119.
[22] CHAO D, KAILI L, JING Z, et al. Collaborative filtering recommendation algorithm classification and comparative study[C]//Proceedings of the 2019 4th International Con-ference on Distance Education and Learning, Shanghai, May 24-27, 2019. New York: ACM, 2019: 106-111.
[23] ELBADRAWY A, KARYPIS G. Domain-aware grade prediction and top-n course recommendation[C]//Proceedings of the 10th ACM Conference on Recommender Systems, Boston, Sep 15-19, 2016. New York: ACM, 2016: 183-190.
[24] YANG D, PIERGALLINI M, HOWLEY I, et al. Forum thread recommendation for massive open online courses[C]// Proceedings of the 7th International Conference on Educa-tional Data Mining, London, Jul 4-7, 2014: 257-260.
[25] ADAMOPOULOS P. On discovering non-obvious recommen-dations: using unexpectedness and neighborhood selection methods in collaborative filtering systems[C]//Proceedings of the 7th ACM International Conference on Web Search and Data Mining. New York: ACM, 2014: 655-660.
[26] SYMEONIDIS P, MALAKOUDIS D. MoocRec.com: massive open online courses recommender system[C]//Proceedings of the Poster Track of the 10th ACM Conference on Recom-mender Systems, Boston, Sep 17, 2016.
[27] KIPF T N, WELLING M. Semi-supervised classification with graph convolutional networks[J]. arXiv:1609.02907, 2016.
[28] GAO H, WANG Z, JI S. Large-scale learnable graph convo-lutional networks[C]//Proceedings of the 24th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, London, Aug 19-23, 2018. New York: ACM, 2018: 1416-1424.
[29] WANG X, JI H, SHI C, et al. Heterogeneous graph atten-tion network[C]//Proceedings of the 2019 World Wide Web Conference, San Francisco, May 13-17, 2019. New York: ACM, 2019: 2022-2032.
[30] SHI C, HU B, ZHAO W X, et al. Heterogeneous information network embedding for recommendation[J]. IEEE Transactions on Knowledge and Data Engineering, 2018, 31(2): 357-370.
[31] YU J, LUO G, XIAO T, et al. MOOCCube: a large-scale data repository for NLP applications in MOOCs[C]//Pro-ceedings of the 58th Annual Meeting of the Association for Computational Linguistics, Jul 5-10, 2020. Stroudsburg: ACL, 2020: 3135-3142.
[32] MIKOLOV T, CHEN K, CORRADO G, et al. Efficient estimation of word representations in vector space[J]. arXiv:1301.3781, 2013.
[33] YING R, HE R, CHEN K, et al. Graph convolutional neural networks for web-scale recommender systems[C]//Proceedings of the 24th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, London, Aug 19-23, 2018. New York: ACM, 2018: 974-983.
[34] PAL A, EKSOMBATCHAI C, ZHOU Y, et al. Pinnersage: multi-modal user embedding framework for recommendations at pinterest[C]//Proceedings of the 26th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, Aug 23-27, 2020. New York: ACM, 2020: 2311-2320.
[35] DONG Y, CHAWLA N V, SWAMI A. Metapath2vec: scalable representation learning for heterogeneous networks[C]//Pro-ceedings of the 23rd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, Halifax, Aug 13-17, 2017. New York: ACM, 2017: 135-144.
[36] WANG S, CHEN Z, LI D, et al. Attentional heterogeneous graph neural network: application to program reidentification[C]//Proceedings of the 2019 SIAM International Conference on Data Mining, Calgary, May 2-4, 2019. Philadelphia: SIAM, 2019: 693-701.
[37] GORI M, MONFARDINI G, SCARSELLI F. A new model for learning in graph domains[C]//Proceedings of the 2005 IEEE International Joint Conference on Neural Networks, Montreal, Jul 31-Aug 4, 2005. Piscataway: IEEE, 2005: 729-734.
[38] ZHAO Z, ZHANG X, ZHOU H, et al. HetNERec: hetero-geneous network embedding based recommendation[J]. Know-ledge-Based Systems, 2020, 204: 106218.
[39] HE X, DENG K, WANG X, et al. LightGCN: simplifying and powering graph convolution network for recommendation[C]//Proceedings of the 43rd International ACM SIGIR Conference on Research and Development in Information Retrieval, Jul 25-30, 2020. New York: ACM, 2020: 639-648.
[40] CHEN T, SUN Y. Task-guided and path-augmented hetero-geneous network embedding for author identification[C]// Proceedings of the 10th ACM International Conference on Web Search and Data Mining, Cambridge, Feb 6-10, 2017. New York: ACM, 2017: 295-304.
[41] GROVER A, LESKOVEC J. node2vec: scalable feature learning for networks[C]//Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. New York: ACM, 2016: 855-864.
[42] RENDLE S, FREUDENTHALER C, GANTNER Z, et al. BPR: Bayesian personalized ranking from implicit feedback[C]//Proceedings of the 25th Conference on Uncertainty in Artificial Intelligence, Montreal, Jun 18-21, 2009. New York: ACM, 2009: 452-461.
[43] WANG X, HE X, WANG M, et al. Neural graph collabora-tive filtering[C]//Proceedings of the 42nd International ACM SIGIR Conference on Research and Development in Infor-mation Retrieval, Paris, Jul 21-25, 2019. New York: ACM, 2019: 165-174.
[44] MAO K, ZHU J, XIAO X, et al. UltraGCN: ultra simplifi-cation of graph convolutional networks for recommendation[C]//Proceedings of the 30th ACM International Conference on Information & Knowledge Management. New York: ACM, 2021: 1253-1262.
[45] WANG M, ZHENG D, YE Z, et al. Deep graph library: a graph-centric, highly-performant package for graph neural networks[J]. arXiv:1909.01315, 2019.