Skip header navigation

Conference Proceeding

A multivariate correlation distance for vector spaces

Connor R & Moss R (2012) A multivariate correlation distance for vector spaces. In: Navarro G & Pestov V (eds.) Similarity Search and Applications: 5th International Conference, SISAP 2012, Toronto, ON, Canada, August 9-10, 2012. Proceedings. Lecture Notes in Computer Science, 7404. Similarity Search and Applications: 5th International Conference, SISAP 2012, Toronto, 09.08.2012-10.08.2012. Berlin, Heidelberg: Springer Verlag, pp. 209-225.

We investigate a distance metric, previously defined for the measurement of structured data, in the more general context of vector spaces. The metric has a basis in information theory and assesses the distance between two vectors in terms of their relative information content. The resulting metric gives an outcome based on the dimensional correlation, rather than magnitude, of the input vectors, in a manner similar to Cosine Distance. In this paper the metric is defined, and assessed, in comparison with Cosine Distance, for its major properties: semantics, properties for use within similarity search, and evaluation efficiency. We find that it is fairly well correlated with Cosine Distance in dense spaces, but its semantics are in some cases preferable. In a sparse space, it significantly outperforms Cosine Distance over TREC data and queries, the only large collection for which we have a human-ratified ground truth. This result is backed up by another experiment over movielens data. In dense Cartesian spaces it has better properties for use with similarity indices than either Cosine or Euclidean Distance. In its definitional form it is very expensive to evaluate for high-dimensional sparse vectors; to counter this, we show an algebraic rewrite which allows its evaluation to be performed more efficiently. Overall, when a multivariate correlation metric is required over positive vectors, SED seems to be a better choice than Cosine Distance in many circumstances.

Distance metric; multivariate correlation; vector space; cosine distance; similarity search;

Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

Author(s)Connor, Richard; Moss, Robert
Title of seriesLecture Notes in Computer Science
Number in series7404
Publication date31/12/2012
PublisherSpringer Verlag
Place of publicationBerlin, Heidelberg
ISSN of series0302-9743
ConferenceSimilarity Search and Applications: 5th International Conference, SISAP 2012
Conference locationToronto
Scroll back to the top