Adding Full Text Indexing to the Operating System

  • Authors:
  • Kyle Peltonen

  • Affiliations:
  • -

  • Venue:
  • ICDE '97 Proceedings of the Thirteenth International Conference on Data Engineering
  • Year:
  • 1997

Quantified Score

Hi-index 0.00

Visualization

Abstract

Many challenges must be faced when incorporating full text retrieval into the operating system. The search engine must be a nearly invisible, natural extension to the operating system, just like the file system and the network. The search engine must meet user expectations of an operating system, specifically in areas such as performance, fault tolerance, and security. It must handle a very heterogeneous collection of documents, in many formats, many languages and many styles. The search engine must scale with the operating system, from small laptop computers to large multiprocessor servers. The paper is an overview of the challenges faced when incorporating full text indexing into the Microsoft Windows NT/sup TM/ operating system. Specific solutions used by the Microsoft 'Tripoli' search engine, are offered.