Retrieve-Classify-Read: Passage Filtering via Subject Classification for University Question Answering
Why this work is in the frame
A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.
Bibliographic record
Abstract
University FAQs are useful for prospective and current students but the information that they contain is often limited and fails to address specialized questions that students might have. A dedicated Question Answering (QA) model which could answer these specialized university-related questions would be of great use for information accessibility. Prior research has shown that QA systems using reader models that incorporate information from multiple passages have benefited in terms of accuracy and efficiency from carefully selecting optimal subsets of retrieved passages. This study explores the effectiveness of passage filtering via subject classification for an Acadia University QA model by using an oracle filtering model between retriever and reader models. This oracle model filters retrieved passages by removing all those that do not share a subject with the given question. The performance of the QA model that employs filtering is compared to that of a baseline QA model which does not implement passage filtering. Results show that using a passage filtering model reduces the number of passages that get passed to the reader model on average, and that oracle based subject filtering boosts Exact Match and F1 accuracy compared to an unfiltered model.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.002 |
| Science and technology studies | 0.001 | 0.000 |
| Scholarly communication | 0.000 | 0.002 |
| Open science | 0.002 | 0.001 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it