[1]陈帜 张文德.基于多模态数据的学习资源序列推荐研究[J].信息化理论与实践,2021,(01):133-139.
　Research on Learning Resource Sequence Recommendation Based on Multimodal Data Chen Zhi12 Zhang Wende1[J].Information Theory and Practice,2021,(01):133-139.
点击复制

基于多模态数据的学习资源序列推荐研究()
分享到：

《信息化理论与实践》[ISSN:2520-5862/CN:]

卷:
期数:
2021年01
页码:
133-139
栏目:
出版日期:
2021-12-31

文章信息/Info

Title:
Research on Learning Resource Sequence Recommendation Based on Multimodal Data Chen Zhi12 Zhang Wende1
作者:
陈帜 张文德
福州大学信息管理研究所
Author(s):
1
(Institute of Information Management, Fuzhou University, Fuzhou 350108, China) 2 ( Fuzhou University Library, Fuzhou 350108, China)
关键词:
多模态数据 自注意力 序列推荐
Keywords:
Multimodal data Self attention Sequential recommendation
摘要:
目的]以学习者的学习行为序列为研究对象，研究利用多模态数据感知和挖掘学生学习偏好的方法。[方法]利用自注意力机制，挖掘学习者学习行为序列中不同学习资源占据的重要性，根据最近的几次学习行为，生成学习者下次学习最有可能选择的学习资源作为推荐。[结果]在中国大学MOOC平台的真实学习行为数据集上进行验证，所提出的推荐模型比基线方法在具体评测指标上取得了更好的效果。[局限]仅挖掘学习者学习行为和时间数据蕴含的信息，未能考虑设备类型、位置和行为类型等边信息反映的学习偏好。[结论]自注意力机制能够有效分析学习者行为序列，通过近期的学习行为生成个性化的序列推荐结果。
Abstract:
Objective] Taking learners’ behavior sequence as the research object, to study the method of perceiving and mining students’ learning preference using multimodal data. [Methods] Using the self-attention mechanism to mine the importance of different learning resources in the learner’s learning behavior sequence, and based on the recent several learning behavior, generate the learning resources that the learner is most likely to choose for the next learning as a recommendation. [Results] Validated on the real learning behavior dataset of Chinese university MOOC platform, the proposed recommendation model achieves better results on specific evaluation indicators than the baseline methods. [Limitations] Only the information contained in the learner’s learning behavior and time data is mined, and the learning preferences reflected by the side information such as device type, location, and behavior type are not considered. [Conclusions] The self-attention mechanism can effectively analyze the learner’s behavior sequence, and generate personalized sequence recommendation results through the recent learning behavior

参考文献/References:

[1] 李晓明, 张绒. 慕课:理想性、现实性及其对高等教育的潜在影响[J]. 电化教育研究, 2017,38(2): 62-65.

[2] Chango W, Cerezo R, Romero C. Multi-source and multimodal data fusion for predicting academic performance in blended learning university courses[J]. Computers & Electrical Engineering, 2021, 89: 106908.

[3] Ricci F, Rokach L, Shapira B. Introduction to recommender systems handbook[M]//Recommender systems handbook. Springer, Boston, MA, 2011: 1-35.

[4] Lonjarret C, Auburtin R, Robardet C, et al. Sequential recommendation with metric models based on frequent sequences[J]. Data Mining and Knowledge Discovery, 2021, 35(3): 1087-1133.

[5] 程岩. 在线学习中基于群体智能的学习路径推荐方法[J]. 系统管理学报, 2011,20(2): 232-237.

[6] 吴雷, 方卿. 基于改进粒子群算法的学习路径优化方法[J]. 系统科学与数学, 2016,36(12): 2272-2281.

[7] 桂忠艳, 张艳明, 李巍巍. 基于行为序列分析的学习资源推荐算法研究[J]. 计算机应用研究, 2020,37(7): 1979-1982

[8] Mnih V, Heess N, Graves A. Recurrent models of visual attention[C]// Proceedings of the 27th International Conference on Neural Information Processing Systems(NIPS2014), 2014: 2204-2212

[9] Vaswani A, Shazeer N, Parmar N, et al. Attention is all you need[C]// Proceedings of the 31st International Conference on Neural Information Processing Systems (NIPS2017), 2017: 6000-6010.

[10] 张青博, 王斌, 崔宁宁, 等. 基于注意力机制的规范化矩阵分解推荐算法[J]. 软件学报, 2020,31(3): 778-793.

[11] 沈学利, 杜志伟. 融合自注意力机制与长短期偏好的序列推荐模型[J]. 计算机应用研究, 2021,38(5): 1371-1375.

[12] Rendle S, Freudenthaler C, Schmidt-Thieme L. Factorizing personalized markov chains for next-basket recommendation[C]//Proceedings of the 19th international conference on World wide web(WWW2010), 2010: 811-820.

[13] Hidasi B, Karatzoglou A, Baltrunas L, et al. Session-based recommendations with recurrent neural networks[J/OL]. arXiv preprint arXiv:1511.06939, 2015. https://arxiv.org/abs/1511.06939

更新日期/Last Update: 2022-12-20