Novel Problems and Challenges in Language-based Conversational Recommender Systems
Bibliographic record
Abstract
Language-based Conversational Recommender Systems (CRSs) have attracted growing attention as they allow users to express and interactively refine their preferences in natural language. However, there exist open problems in CRSs relating to the challenges users face when articulating accurate natural language preferences and the underlying technologies required to facilitate language-based interactions. Our first contribution addresses the challenge that users have trouble specifying preferences with the right level of specificity. We propose a clarification-based extension of a critiquing-based interaction workflow for CRSs that outperforms state-of-the-art models. In our second contribution, we explore the novel issue of unintended bias in language model-driven conversational recommendation by proposing novel bias evaluation metrics and performing source of bias analysis. In summary, this thesis investigates novel and important challenges in the deployment of language that can help users express their preferences more accurately and allow the identification and resolution of bias that arises in language-model-driven CRSs.
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.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.040 | 0.123 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.003 | 0.002 |
| Bibliometrics | 0.001 | 0.002 |
| Science and technology studies | 0.003 | 0.005 |
| Scholarly communication | 0.007 | 0.014 |
| Open science | 0.003 | 0.007 |
| Research integrity | 0.004 | 0.007 |
| Insufficient payload (model declined to judge) | 0.002 | 0.001 |
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 itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".