Interval query indexing for efficient stream processing

  • Authors:
  • Kun-Lung Wu;Shyh-Kwei Chen;Philip S. Yu

  • Affiliations:
  • IBM T.J. Watson Research Center, Hawthorne, NY;IBM T.J. Watson Research Center, Hawthorne, NY;IBM T.J. Watson Research Center, Hawthorne, NY

  • Venue:
  • Proceedings of the thirteenth ACM international conference on Information and knowledge management
  • Year:
  • 2004

Quantified Score

Hi-index 0.00

Visualization

Abstract

A large number of continual range queries can be issued against a data stream. Usually, a main memory-based query index with a small storage cost and a fast search time is needed, especially if the stream is rapid. In this paper, we present a CEI-based query index that meets both criteria for efficient processing of continual interval queries in a streaming environment. This new query index is centered around a set of predefined virtual containment-encoded intervals, or CEIs. The CEIs are used to first decompose query intervals and then perform efficient search operations. The CEIs are defined and labeled such that containment relationships among them are encoded in their IDs. The containment encoding makes decomposition and search operations efficient because integer additions and logical shifts can be used to carry out most of the operations. Simulations are conducted to evaluate the effectiveness of the CEI-based query index and to compare it with alternative approaches. The results show that the CEI-based query index significantly outperforms existing approaches in terms of both storage cost and search time.