In recommender models, disentangled representations correlate with representation interpretability but not with recommendation effectiveness.
IEEE transactions on pattern analysis an d machine intelligence 35(8), 1798–1828 (2013)
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.IR 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Are Representation Disentanglement and Interpretability Linked in Recommendation Models? A Critical Review and Reproducibility Study
In recommender models, disentangled representations correlate with representation interpretability but not with recommendation effectiveness.