On principal component analysis, cosine and Euclidean measures in information retrieval.
Tuomo KoreniusJorma LaurikkalaMartti JuholaPublished in: Inf. Sci. (2007)
Keyphrases
- information retrieval
- principal component analysis
- euclidean distance
- dimensionality reduction
- principal components
- independent component analysis
- cosine measure
- information extraction
- covariance matrix
- learning to rank
- cosine distance
- similarity measure
- information retrieval systems
- low dimensional
- euclidean space
- face recognition
- search engine
- discriminant analysis
- dimension reduction
- evaluation measures
- vector space model
- precision recall
- vector space
- test collection
- document collections
- feature space
- text mining
- basis functions
- singular value decomposition
- retrieval systems
- information seeking
- feature vectors
- information filtering
- tf idf
- linear discriminate analysis
- quantitative measures
- lower dimensional
- language model
- information access
- query expansion
- retrieval effectiveness