Cache and Priority Queue Based Approximation Technique for a Stream of Similarity Search Queries

Investor logo

Warning

This publication doesn't include Faculty of Economics and Administration. It includes Faculty of Informatics. Official publication website can be found on muni.cz.
Authors

NÁLEPA Filip BATKO Michal ZEZULA Pavel

Year of publication 2017
Type Article in Proceedings
Conference Similarity Search and Applications : 10th International Conference, SISAP 2017, Munich, Germany, October 4-6, 2017, Proceedings
MU Faculty or unit

Faculty of Informatics

Citation
Doi http://dx.doi.org/10.1007/978-3-319-68474-1_2
Field Informatics
Keywords approximate similarity search; stream of kNN queries
Description Content-based similarity search techniques have been employed in a variety of today applications. In our work, we aim at the scenario when the similarity search is applied in the context of stream processing. In particular, there is a stream of query objects which need to be evaluated. Our goal is to be able to cope with the rate of incoming query objects (i.e., to reach sufficient throughput) and, at the same time, to preserve the quality of the obtained results at high levels. We propose an approximation technique for the similarity search which combines the probability of an indexed object to be a part of a query result and the time needed to examine the object. We are able to achieve better trade-off between the efficiency (processing time) and the quality (precision) of the similarity search compared to traditional priority queue based approximation techniques.
Related projects:

You are running an old browser version. We recommend updating your browser to its latest version.