Use a cohort-wide instructor dashboard of topic-level question volume, with assistant-surfaced prerequisite review suggestions as an extension¶

Strategy · All strategies
Evidence · no claims cited

Description¶

The article's envisioned practical use of the pipeline: converting passively logged assistant interactions into topic-level difficulty signals without additional student burden. The authors state "The primary use we envision is a cohort-wide instructor dashboard", with the same signal passed to the conversational assistant as suggestions for prerequisite review when a student asks about a downstream topic, or surfaced through the LMS as student-specific recommendations. Instructors can verify the ranking against independent survey evidence.

Design Implications¶

Context¶

Requirements¶

• A deployed conversational AI teaching assistant producing logged question events, and a curriculum knowledge graph grounding the topic labels

Constraints¶

• The H2 result is aggregate at the topic level and does not by itself identify which individual students are struggling on which topic

Target Learners¶

• students in large online courses using conversational AI teaching assistants

Target Learning Goals¶

• identifying topics warranting instructional attention
• targeting prerequisite review

Related Strategies¶

• Use support-level filtering across cohorts to separate persistent from cohort-specific challenging topics

Examples¶

-

Key Sources¶

• Youssef Medhat, Junsoo Park, Ploy Thajchayapong and Ashok K. Goel. (2026). Detecting Knowledge Gaps from Conversational AI Interactions Using Curriculum Prerequisite Graphs. CSEDM'26: 10th Workshop on Computer Science Educational Data Mining. https://arxiv.org/abs/2606.10736