Stop Playing the Guessing Game! Evaluating Conversational Recommender Systems via Target-free User Simulation
Bibliographic record
Abstract
Recent developments in Conversational Recommender Systems (CRSs) have focused on simulating real-world interactions between users and CRSs to create more realistic evaluation environments.Despite considerable advancements, reliably assessing the capability of CRSs in eliciting user preferences remains a significant challenge.We observe that user-CRS interactions in existing evaluation protocols resemble a guessing game, as they construct targetbiased simulators pre-encoded with target item knowledge, thereby allowing the CRS to shortcut the elicitation process.Moreover, we reveal that current evaluation metrics, which predominantly emphasize single-turn recall of target items, suffer from target ambiguity in multiturn settings and overlook the intermediate process of preference elicitation.To address these issues, we introduce PEPPER, a novel CRS evaluation protocol with target-free user simulators that enable users to gradually discover their preferences through enriched interactions, along with detailed measures for comprehensively assessing the preference elicitation capabilities of CRSs.Through extensive experiments, we validate PEPPER as a reliable evaluation protocol and offer a thorough analysis of how effectively current CRSs perform in preference elicitation and recommendation.
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 | 0.010 | 0.041 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.000 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.001 | 0.002 |
| Open science | 0.001 | 0.002 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.003 | 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".