The R*-tree: an efficient and robust access method for points and rectangles
SIGMOD '90 Proceedings of the 1990 ACM SIGMOD international conference on Management of data
Combining fuzzy information from multiple systems (extended abstract)
PODS '96 Proceedings of the fifteenth ACM SIGACT-SIGMOD-SIGART symposium on Principles of database systems
Querying multimedia data from multiple repositories by content: the Garlic project
Proceedings of the third IFIP WG2.6 working conference on Visual database systems 3 (VDB-3)
Optimal aggregation algorithms for middleware
PODS '01 Proceedings of the twentieth ACM SIGMOD-SIGACT-SIGART symposium on Principles of database systems
Reminiscences on Influential Papers
ACM SIGMOD Record
Supporting Ranked Boolean Similarity Queries in MARS
IEEE Transactions on Knowledge and Data Engineering
Query Processing Issues in Image(Multimedia) Databases
ICDE '99 Proceedings of the 15th International Conference on Data Engineering
Boolean + ranking: querying a database by k-constrained optimization
Proceedings of the 2006 ACM SIGMOD international conference on Management of data
| Hi-index | 0.00 |
This paper deals with the issue of combining fuzzy results obtained from two individual systems in multimedia query processing. Consider a query such as retrieve images similar to I1 by color AND with associated text flower. Suppose we have a color subsystem which can return a sorted list of images based on similarity with I1 and a text subsystem which can return a sorted list of images based on similarity with flower. Our task is to combine these two results. In other words, we need to evaluate the fuzzy combining function AND, giving a sorted list of the images.Fagin has proved that the probabilistic complexity of the problem is almost linear (Fagin 1996), and our multi-step algorithm (Nepal & M. V. Ramakrishna 1999) is an optimal uniform algorithm (Fagin, Lotem & Naor 2001). In view of this inherent limitation, we investigated a non-uniform heuristic approach. In this paper, we discuss the problems of processing such queries, also referred to as aggregation queires in multimedia databases. The experimental results presented show that our "minimum depth first search" heuristic approach out-performs other uniform algorithms in general and by over 90% when the distribution of similarity values is not uniform.