arXiv：cs.LG（机器学习，全量分类）· Hitoshi Inoue, Koichi Yasutake·· 15 小时前AI 评分28
异步在线学习中的参与敏感收敛与"先碎片化后收敛"模式：跨 22 门 OULAD 课程的拓扑分析
Participation-Sensitive Convergence and the Fragment First, Converge Later Pattern in Asynchronous Online Learning: A Topological Analysis Across 22 OULAD Courses
AI 导读
研究基于 Zigzag 持续同调的 $\beta_0$ 指标，对全部 22 门 OULAD 课程（N>22,000，857 个周对）进行分析，发现 $\beta_0$ 变化与活跃学习者数量变化高度共变（合并 r=0.387，20/22 门课程中位 r_delta=0.459），表明 $\beta_0$ 是参与敏感的结构指标。
正文
Abstract:Asynchronous online learning offers temporal flexibility at a structural cost: learning communities tend to fragment rather than cohere. $\beta_0$, the number of disconnected behavioral clusters from Zigzag Persistent Homology, serves as a cohort-level indicator of this structure. Two questions remained unverified at scale: (1) does apparent $\beta_0$ convergence reflect genuine behavioral alignment or learner dropout? and (2) do assessment deadlines produce reproducible fragmentation-convergence cycles? We address both across all 22 OULAD courses (N > 22,000; 857 week-pairs). Changes in $\beta_0$ strongly co-vary with active learner changes (pooled r = 0.387; median per-course r_delta = 0.459, 20/22 courses), identifying $\beta_0$ as a participation-sensitive indicator: $\beta_0$ and active learner counts co-respond to deadline events rather than one causing the other. Deadlines produced fragmentation in 82.6% of assessments and the full Fragment First, Converge Later (FFCL) cycle in 60.2%. 3-phase analysis confirmed structural fragmentation as the dominant long-term trajectory (90.9% of courses), moderated by curriculum structure. These findings establish $\beta_0$ as a participation-sensitive structural indicator with direct implications for AI-augmented learning analytics design.
| Comments: | Author's version, posted under the non-commercial rights retained in the APSCE copyright transfer agreement |
| Subjects: | Computers and Society (cs.CY); Machine Learning (cs.LG) |
| Cite as: | arXiv:2610.01738 [cs.CY] |
| (or arXiv:2610.01738v1 [cs.CY] for this version) | |
| https://doi.org/10.48550/arXiv.2610.01738 arXiv-issued DOI via DataCite (pending registration) |
|
| Journal reference: | Proceedings of ICLEA 2026: 2nd International Conference on Learning Evidence and Analytics, Asia-Pacific Society for Computers in Education, 2026 |
Submission history
From: Hitoshi Inoue [view email]
[v1]
Thu, 1 Oct 2026 14:10:36 UTC (340 KB)
来源：arXiv：cs.LG（机器学习，全量分类） · arxiv.org