Similarity indexing by means of a metric

This paper presents a method for indexing a large data set by means of a metric and indicates its use for quantified proximity searching (search precision is a parameter of the query). We make a proposal for adding the property of a dimension to a metric and show that this is compatible to our customized understanding of a dimension. We present an algorithm which computes, in optimal time, an index on the data set which makes full use of this dimension. The index can be regarded as a materialized view for supporting similarity queries with predictable performance. The design of the query algorithms are robust with respect to skewed data and the method can be applied in a distributed C/S-environment (such as WWW/cgi or SQLnet).