Processing complex similarity queries with distance-based access methods

Abstract

Abstract. Efficient evaluation of similarity queries is one of the basic requirements for advanced multimedia applications. In this paper, we consider the relevant case where complex similarity queries are defined through a generic language L and whose predicates refer to a single feature F . Contrary to the language level which deals only with similarity scores, the proposed evaluation process is based on distances between feature values -known spatial or metric indexes use distances to evaluate predicates. The proposed solution suggests that the index should process complex queries as a whole, thus evaluating multiple similarity predicates at a time. The flexibility of our approach is demonstrated by considering three different similarity languages, and showing how the M-tree access method has been extended to this purpose. Experimental results clearly show that performance of the extended M-tree is consistently better than that of state-of-the-art search algorithms.

PAC nearest neighbor queries: Approximate and controlled search in high-dimensional and metric spaces

“…Another interesting research issue would be to apply our results to the case of complex NN queries, where more than one similarity criterion has to be applied in order to determine the overall similarity of two objects [14,12].…”

Combining Approaches to Information Retrieval

“…Lee's experiments do provide evidence that these combination operators perform significantly worse than summing. Ciaccia et al, 1998, also discuss ranking in a multimedia database environment. They are concerned primarily with the efficiency of combination, as is Fagin, and present performance results for a range of combination operators.…”

The Basic Principles of Metric Indexing

“…However, that number can be high -in many cases leading to unrealistically high space requirements. 24 One possible tradeoff is to reduce the number of pivots below the optimal. As noted in Sect.…”

“…• Methods dealing with query types other than range search, * such as the similarity self-join (nearest pair queries) of the eD-index [19], the multimetric or complex searches of M 3 -tree [20] and others [21][22][23][24], searching with user-defined metrics, as in the QIC-M-tree [25], or the incremental search of Hjaltason and Samet [26].…”