1376-GRM A Verifiable LLM Pipeline for Translating Educational Research into Instructor-Ready Teaching Tactics

Presenter Information

Location

https://campus.kennesaw.edu/colleges-departments/ccse/events/computing-showcase/fa26-cday-program.php

Document Type

Event

Start Date

20-11-2026 2:00 PM

Description

General AI chatbots answer teaching questions but rarely show checkable evidence. We present PDPRS, a pipeline that turns teaching research into Pedagogical Design Patterns (a classroom problem, a research-supported solution, its rationale and a common pitfall) and a Teaching Coach that answers from them. Using only local open models, it screened 3,245,254 records, kept 15,510 verified claims from 1,354 full-text papers and wrote 178 patterns, verifying each output before the next stage uses it. On held-out papers, 99% of extracted claims were correct and 91% of findings were found; 94% of the coach's answer sentences were supported by the study cited

This document is currently not available here.

Share

COinS
 
Nov 20th, 2:00 PM

1376-GRM A Verifiable LLM Pipeline for Translating Educational Research into Instructor-Ready Teaching Tactics

https://campus.kennesaw.edu/colleges-departments/ccse/events/computing-showcase/fa26-cday-program.php

General AI chatbots answer teaching questions but rarely show checkable evidence. We present PDPRS, a pipeline that turns teaching research into Pedagogical Design Patterns (a classroom problem, a research-supported solution, its rationale and a common pitfall) and a Teaching Coach that answers from them. Using only local open models, it screened 3,245,254 records, kept 15,510 verified claims from 1,354 full-text papers and wrote 178 patterns, verifying each output before the next stage uses it. On held-out papers, 99% of extracted claims were correct and 91% of findings were found; 94% of the coach's answer sentences were supported by the study cited