Authors:
Naoya Higuchi
1
;
Yasunobu Imamura
1
;
Tetsuji Kuboyama
2
;
Kouichi Hirata
1
and
Takeshi Shinohara
1
Affiliations:
1
Kyushu Institute of Technology, Japan
;
2
Gakushuin University, Japan
Keyword(s):
Similarity Search, Sketches, Ball Partitioning, Hamming Distance, Dimension Reduction, Distance Lower Bound, Quantized Images.
Related
Ontology
Subjects/Areas/Topics:
Applications
;
Data Engineering
;
Ensemble Methods
;
Information Retrieval
;
Ontologies and the Semantic Web
;
Pattern Recognition
;
Similarity and Distance Learning
;
Software Engineering
;
Theory and Methods
Abstract:
In this paper, we discuss sketches based on ball partitioning (BP), which are compact bit sequences representing multidimensional data. The conventional nearest search using sketches consists of two stages. The
first stage selects candidates depending on the Hamming distances between sketches. Then, the second stage
selects the nearest neighbor from the candidates. Since the Hamming distance cannot completely reflect the
original distance, more candidates are needed to achieve higher accuracy. On the other hand, we can regard
BP sketches as quantized images of a dimension reduction. Although quantization error is very large if we use
only sketches to compute distances, we can partly recover distance information using query. That is, we can
compute a lower bound of distance between a query and a data using only query and the sketch of the data.
We propose candidate selection methods at the first stage using the lower bounds. Using the proposed method,
higher level of accuracy
for nearest neighbor search is shown through experimenting on multidimensional data
such as images, music and colors.
(More)