MétaCan
Menu
Back to cohort
Record W2400248106 · doi:10.1145/2724720

Improving Top-N Recommendation for Cold-Start Users via Cross-Domain Information

2015· article· en· W2400248106 on OpenAlex

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.

affAt least one author lists a Canadian institution in the pinned OpenAlex snapshot.

Bibliographic record

VenueACM Transactions on Knowledge Discovery from Data · 2015
Typearticle
Languageen
FieldComputer Science
TopicRecommender Systems and Techniques
Canadian institutionsWestern University
Fundersnot available
KeywordsCold start (automotive)Computer scienceDomain (mathematical analysis)Matrix decompositionCluster analysisPrecision and recallRecommender systemData miningArtificial intelligenceMachine learningInformation retrievalMathematicsEngineering

Abstract

fetched live from OpenAlex

Making accurate recommendations for cold-start users is a challenging yet important problem in recommendation systems. Including more information from other domains is a natural solution to improve the recommendations. However, most previous work in cross-domain recommendations has focused on improving prediction accuracy with several severe limitations. In this article, we extend our previous work on clustering-based matrix factorization in single domains into cross domains. In addition, we utilize recent results on unobserved ratings. Our new method can more effectively utilize data from auxiliary domains to achieve better recommendations, especially for cold-start users. For example, our method improves the recall to 21% on average for cold-start users, whereas previous methods result in only 15% recall in the cross-domain Amazon dataset. We also observe almost the same improvements in the Epinions dataset. Considering that it is often difficult to make even a small improvement in recommendations, for cold- start users in particular, our result is quite significant.

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 imitation

Not 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.

metaresearch head score (Codex)0.001
metaresearch head score (Gemma)0.000
Version: codex-gemma-dda1882f352aValidation status: machine_predicted_unvalidated
Candidate categoriesnone
Consensus categoriesnone
DomainCandidate signal: none · Consensus signal: none
Study designCandidate signal: Other design · Consensus signal: none
GenreCandidate signal: Methods · Consensus signal: none
Teacher disagreement score0.938
Threshold uncertainty score0.952

Codex and Gemma teacher scores by category

CategoryCodexGemma
Metaresearch0.0010.000
Meta-epidemiology (narrow)0.0000.000
Meta-epidemiology (broad)0.0000.000
Bibliometrics0.0000.000
Science and technology studies0.0000.000
Scholarly communication0.0010.012
Open science0.0020.000
Research integrity0.0000.000
Insufficient payload (model declined to judge)0.0000.000

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.

Opus teacher head0.070
GPT teacher head0.320
Teacher spread0.250 · how far apart the two teachers sit on this one work
Validation statusscore_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it