SiLi Index: Data Structure for Fast Vector Space Searching
Autoři | |
---|---|
Rok publikování | 2019 |
Druh | Článek ve sborníku |
Konference | Proceedings of the Thirteenth Workshop on Recent Advances in Slavonic Natural Languages Processing, RASLAN 2019 |
Fakulta / Pracoviště MU | |
Citace | HERMAN, Ondřej a Pavel RYCHLÝ. SiLi Index: Data Structure for Fast Vector Space Searching. In Horák, Aleš and Rychlý, Pavel and Rambousek, Adam. Proceedings of the Thirteenth Workshop on Recent Advances in Slavonic Natural Languages Processing, RASLAN 2019. Brno: Tribun EU, 2019, s. 111-116. ISBN 978-80-263-1530-8. |
www | https://nlp.fi.muni.cz/raslan/2019/paper07-herman.pdf |
Klíčová slova | word embeddings; vector space; semantic similarity |
Popis | Nearest neighbor queries in high-dimensional spaces are ex-pensive. In this article, we propose a method of building and querying astand-alone data structure, SiLi (SimilarityList) Index, which supports ap-proximating the results of k-NN queries in high-dimensional spaces, whileusing a significantly reduced amount of system memory and processortime compared to the usual brute-force search methods. |
Související projekty: |