Abstract
Our work shows that the query latency for selective search over a topically partitioned collection can be reduced by up to 55%. We achieve this by physically storing the documents in each topical cluster across all shards and building a cluster-skipping index at each shard. Our approach also achieves uniform load balance among the shards.
Cite
CITATION STYLE
APA
Hafizoglu, F., Kucukoglu, E. C., & Altingovde, I. S. (2017). On the efficiency of selective search. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 10193 LNCS, pp. 705–712). Springer Verlag. https://doi.org/10.1007/978-3-319-56608-5_69
Register to see more suggestions
Mendeley helps you to discover research relevant for your work.
Already have an account? Sign in
Sign up for free