An efficient algorithm for hyperspherical range query processing in high-dimensional data space

摘要

In this paper, we proposed the SPY-TEC, a new in- dexing technique for similarity search which is very frequently used in applications such as content-based multimedia retrieval. We also showed that the SPY- TEC outperforms other related techniques including The Pyramid-Technique when processing hyperspheri- Cal range queries in high-dimensional data space. For highly skewed data distributions or queries, the SPY-TEC may perform worse than other index struc- Tures.