Universite de Montreal at TREC 2013: Experiments with Quantum Language Models in the Web Track.
Bibliographic record
Abstract
In TREC 2013, we focus on addressing the challenges posed by the Web track using our recently proposed Quantum Language Modeling (QLM) approach for IR [1]. QLM can be considered as a dependence model for IR for its capability of representing and integrating compound term dependencies into the scoring function. Among the main properties of the model, two of them make it stand out from the literature of existing dependence models (such as MRF [3]). First, QLM does not combine scores obtained from matching single terms and from matching compound dependencies, which makes it virtually parameterless. This is quite an appealing property for an IR system, especially when a new dataset such as ClueWeb12 is released and no previous training examples can be leveraged to fine-tune important parameters. The second peculiar feature of our model is its ability to automatically fallback onto the baseline bag-ofwords score in the case that the required dependence relationship does not hold in the document. This is expected to bring improved robustness w.r.t. the baseline ranking. In the light of these considerations, the Web Track ad-hoc and robustness task seem the perfect testbeds for our model. In what follows we briefly review some of the theoretical background of QLM before delving into the description of the submitted runs and obtained results.
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 | 0.012 | 0.021 |
| Meta-epidemiology (narrow) | 0.003 | 0.001 |
| Meta-epidemiology (broad) | 0.002 | 0.001 |
| Bibliometrics | 0.001 | 0.002 |
| Science and technology studies | 0.003 | 0.001 |
| Scholarly communication | 0.002 | 0.004 |
| Open science | 0.003 | 0.002 |
| Research integrity | 0.003 | 0.004 |
| Insufficient payload (model declined to judge) | 0.011 | 0.006 |
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".