Database-inspired search

  • Authors:
  • David Konopnicki;Oded Shmueli

  • Affiliations:
  • IBM Haifa Research Laboratory, Haifa, Israel;Technion, Haifa, Israel

  • Venue:
  • VLDB '05 Proceedings of the 31st international conference on Very large data bases
  • Year:
  • 2005

Quantified Score

Hi-index 0.00

Visualization

Abstract

"W3QL: A Query Language for the WWW", published in 1995, presented a language with several distinctive features. Employing existing indexes as access paths, it allowed the selection of documents using conditions on semi-structured documents and maintaining dynamic views of navigational queries. W3QL was capable of automatically filling out forms and navigating through them. Finally, in the SQL tradition, it was a declarative query language, that could be the subject of optimization.Ten years later, we examine some current trends in the domain of search, namely the emergence of system-level search services and of the semantic web. In this context, we explore whether W3QL's ideas are still relevant to help improve information search and retrieval. We identify two main environments for searching, the enterprise and the web at large. Both environments could benefit from database-inspired integration language, and an execution system that implements it.