On the robustness and discriminative power of information retrieval metrics for top-N recommendation.
Daniel ValcarceAlejandro BellogínJavier ParaparPablo CastellsPublished in: RecSys (2018)
Keyphrases
- discriminative power
- information retrieval
- computational cost
- invariant features
- bag of words
- information retrieval systems
- visual words
- sparse coding
- text mining
- ranked list
- document collections
- search engine
- retrieval systems
- pairwise
- image classification
- document retrieval
- machine learning
- palmprint
- retrieval model
- relevant documents
- co occurrence
- knn
- multiscale
- face recognition
- image processing