Summary: E∆cient Object-Relational Interval Management
and Beyond ?
Lars Arge and Andrew Chatham
Department of Computer Science, Duke University, Durham, NC 27708, USA.
Abstract. Recently, the object-relational access method paradigm|the
idea of designing index structures that can be built on top of the SQL
layer of any relational database server|was proposed as a way to design
easy to implement indexes while obtaining strong robustness, perfor-
mance, and integration into transaction management for free.
In this paper, we describe an object-relational index for the 3-sided range
indexing problem. Previously an object-relational index was only known
for the interval management problem, which is a special case of the 3-
sided range indexing problem. Our new index is e∆cient in the worst-
case, and it can be used to answer all general interval relationship queries
e∆ciently. The previously known index were only able to answer 7 out of
13 possible relationship queries e∆ciently. We also describe a (limited)
experimental study of a simplied version of our structure.
In order to be successful, any data model in a large database requires e∆cient