S. Nepal and M. V. Ramakrishna. Query processing issues in image (multimedia) databases. In ICDE 1999.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....no work has been published on the problem that MERGE tackles, i.e. merging the union of the ranked results from multiple sources. However many papers have been published on answering ranked queries when the information about each object is distributed across multiple sources. In [12] 18] 17] [23], algorithms are provided to combine ranked lists of attributes in order to e#ciently retrieve the top results according to an aggregate function of the attributes. In these papers a sorted list is used for each attribute in order to e#ciently retrieve the top N ranked results from a single ....

S. Nepal and M. Ramakrishna. Query processing issues in image (multimedia) databases. ICDE, 1999.

.... that it may appear straightforward to transform an expensive predicate into a normal one: By probing every object for its score, one can build a search index for the predicate (to access objects scored above a threshold or in the sorted order) as required by the current processing frameworks [4, 5, 6, 7, 8, 9, 10] (Section 3.3) This naive approach requires a sequential scan, or complete probing, over the entire database: A database of objects will need sequential probes for each expensive predicate. Such complete probing at query time is clearly unacceptable in most cases. This paper addresses ....

....Microsoft SQL Server, IBM DB2, Oracle, and PostgreSQL) support such predicates. Top k queries have been developed recently in two different contexts. First, in a middleware environment, Fagin [7, 8] pioneered ranked queries and established the well known algorithm (with its improvements in [9, 10]) 12] generalizes to handling arbitrary joins as combining constraints. As Section 3 discusses, these works assume sorted access of search predicates. This paper thus studies probe predicates without efficient sorted access. Secondly, ranked queries were also proposed as a layer on top of ....

[Article contains additional citation context not shown here]

S. Nepal and M. Ramakrishna. Query processing issues in image(multimedia) databases. ICDE 1999.

....This work was originally motivated by queries to multimedia databases, e.g. to retrieve images. Stated in IR terms, the algorithms also assume that postings in the inverted lists are sorted by their contributions and are accessed in sorted order. However, several of the algorithms proposed in [15, 18, 19, 30, 39] also assume that once a document is encountered in one of the inverted lists, we can efficiently compute its complete score by performing lookups into the other inverted lists. This gives much better pruning than sorted access alone, but in a search engine context it may not be efficient as it ....

S. Nepal and M. Ramakrishna. Query processing issues in image (multimedia) databases. In Proc. of the 15th Annual Int. Conf. on Data Engineering, pages 22--29, March 1999.

....feature indexes must wait in a temporary file or buffer until the completion of the probing and sorting process. Note that both approaches are incremental in nature and can support the get more feature efficiently. Several other optimizations of the above algorithms have been proposed recently [73, 43]. An alternative approach to evaluating top k queries has been proposed by Chaudhuri and Gravano [24, 25] It uses the results in [31] to convert top k queries to alpha cut queries and processes them as filter conditions. Under certain conditions (uniquely graded repository) this approach is ....

Surya Nepal and M.V. Ramakrishna. Query processing issues in image (multimedia) databases. In Proc. 15th Int. Conf. on Data Engineering, pages 22--29, 1999.

....the system deteriorates rapidly as the size of the database increases. Many algorithms have been proposed in the litera ture to address aggregation ranking, including Fagin s algorithm [6] the TA, CA and NRA algorithms [7] the Quick Combine algorithm [9] the multi step aggregation algorithm [15] and the Stream Combine algorithm [10] Some prototype systems have incorporated these algorithms to answer aggregated rank queries, such as the IBM GARLIC middleware [18] which uses Fagin s algorithm. A simulation of Fagin s algorithm is used in [4] as a filter condition for querying ....

....such as the IBM GARLIC middleware [18] which uses Fagin s algorithm. A simulation of Fagin s algorithm is used in [4] as a filter condition for querying multimedia repositories; the Quick Combine and multi step algorithms are used in multimedia retrieval for answering multi feature queries [9, 15]; and the Stream Combine algorithm is used in middleware for heterogeneous environments [10] Recently, Natsev et ah introduced the J algorithm, an incremental algorithm to join ranked inputs based on the A search algorithm. In the rest of this paper we refer to these algorithms as rank join ....

[Article contains additional citation context not shown here]

Surya Nepal and M. V. Ramakrishna. Query processing issues in image (multimedia) databases. In ICDE'99, Sydney, Austrialia, pages 22-29. IEEE Computer Society, 1999.

....no random accesses that finds the top k answers. The problems with FA run deeper than simply the fact that it may perform badly when the aggregation function is not strict. Even when the aggregation function is strict, there may be some databases where FA is much too conservative. Several groups [NR99, GBK00, FLN02] independently found a new algorithm, which is called the threshold algorithm in [FLN02] As we shall discuss, the threshold algorithm is essentially optimal (up to a constant factor) over every database. Amnon Lotem first defined TA, and did extensive simulations comparing it to ....

....Park, in the Fall of 1997. A few years later, Michael Franklin brought the existence of this work to the attention of the This algorithm first appeared in the PODS 96 conference version of [Fag99] 4 author. This led to a collaboration that produced the paper [FLN02] Nepal and Ramakrishna [NR99] were the first to publish an algorithm that is equivalent to TA. They essentially restricted attention to the situation where the aggregation function is the min (for more details on their restriction, see [FLN02] Guntzer, Balke, and Kiessling [GBK00] also define an algorithm that is ....

S. Nepal and M. V. Ramakrishna. Query processing issues in image (multimedia) databases. In Proc. 15th International Conference on Data Engineering (ICDE), pages 22--29, March 1999.

....on the aggregation function, every correct algorithm must, with high probability, incur a similar middleware cost in the worst case. We shall present the threshold algorithm , or TA. This algorithm was discovered independently by (at least) three groups, including Nepal and Ramakrishna [NR99] who were the first to publish) Guntzer, Balke, and Kiessling [GBK00] and ourselves. For more information and comparison, see Section 10 on related work. We shall not discuss the probability model here, including the notion of independence , since it is off track. For details, see ....

....optimal algorithm Lower bound: Lower bound: guesses ok) possible Thm 6.4 2 Thm 9.4 (certain t) No wild TA: m guesses m (t strict) No random NRA: m Thm 8.5 access m Thm 9. 5 (t strict) Table 1: Summary of Upper and Lower Bounds 36 10 Related Work Nepal and Ramakrishna [NR99] define an algorithm that is equivalent to TA. Their notion of optimality is weaker than ours. Further, they make an assumption that is essentially equivalent to the aggregation function being the min. Guntzer, Balke, and Kiessling [GBK00] also define an algorithm that is equivalent to TA. ....

S. Nepal and M. V. Ramakrishna. Query processing issues in image (multimedia) databases. In Proc. 15th International Conference on Data Engineering (ICDE), pages 22--29, March 1999.

....not directly handle sources that provide only a random access interface, which are the focus of our paper. In Section 3.3, however, we adapted Fagin et al. s algorithms to our scenario and experimentally compared the resulting techniques with our new approach in Section 5. Nepal and Ramakrishna [14] and Guntzer et al. 9] presented variations of Fagin s original FA algorithm [6] for processing queries over multimedia databases. In particular, Guntzer et al. 9] reduce the number of random accesses through the introduction of more stop condition tests and by exploiting the data distribution. ....

S. Nepal and M. V. Ramakrishna. Query processing issues in image (multimedia) databases. In Proc. of the 15th International Conference on Data Engineering, 1999.

....(#BL.o,BL.fv,BL.a (# SI.a.time,SI.a. date (SI) ## NN 5 BL) ## NN 1 RES) 3 Related works During the last decade, several systems that support content based query have been proposed (see the review in [1] Some of the commonly known prototypes are systems such as MARS [4] DISIMA [3] and CHITRA [5]. Though many works exist, there are very little of them that consider a multimedia join operation that associates two sets of data for similarity. For example, the MARS system allows complex query formulation by an intelligent query refinement tool for the user interaction, but does not support ....

S. Nepal and M.V. Ramakrishna. Query processing issues in image(multimedia) databases. In Proceedings of the Intl. Conference on Data Engineering (ICDE), pages 22--29, Sydney, Australia, March 1999.

....on the aggregation function, every correct algorithm must, with high probability, incur a similar middleware cost in the worst case. We shall present the threshold algorithm , or TA. This algorithm was discovered independently by (at least) three groups, including Nepal and Ramakrishna [NR99] who were the first to publish) Guntzer, Balke, and Kiessling [GBK00] and ourselves. 4 For more information and comparison, see Section 7 on related work. 3 We shall not discuss the probability model here, including the notion of independence , since it is off track. For details, see ....

....S Thm 6.2 m 2 Thm 6.4 guesses ok) possible (certain t) No wild TA: m (m Gamma1)m 2 c R c S Thm 4.3 guesses m (m Gamma1)m 2 c R c S Thm 6.1 (t strict) No random NRA: m Thm 5.5 access m Thm 6. 5 (t strict) Table 1: Upper and Lower Bounds 7 Related Work Nepal and Ramakrishna [NR99] define an algorithm that is equivalent to TA. Their notion of optimality is weaker than ours. Further, they make an assumption that is essentially equivalent to the aggregation function being the min. 16 Guntzer, Balke, and Kiessling [GBK00] also define an algorithm that is equivalent to TA. ....

S. Nepal and M. V. Ramakrishna. Query processing issues in image (multimedia) databases. In Proc. 15th International Conference on Data Engineering (ICDE), pages 22--29, March 1999.

....along with an assumption on the aggregation function, every correct algorithm must, with high probability, incur a similar middleware cost. We shall present the threshold algorithm , or TA. This algorithm has been defined and studied by (at least) three groups, including Nepal and Ramakrishna [NR99] who were the first to publish) Guntzer, Balke, and Kiessling [GBK00] and ourselves. 3 For more information and comparison, see Section 6 on related work. 2 We shall not discuss the probability model here, including the notion of independence , since it is off track. For details, see ....

....each of the h Gamma 2 objects at the top of each of the three lists. Since we take all of these objects to be distinct, this is 6(h Gamma 2) random accesses. In particular, Theorem 5.7 would be false if we were to replace CA by the intermittent algorithm. 6 Related Work Nepal and Ramakrishna [NR99] define an algorithm that is equivalent to TA. Their notion of optimality is weaker than ours. Further, they make an assumption that is essentially equivalent to the aggregation function being the min. 9 Guntzer, Balke, and Kiessling [GBK00] also define an algorithm that is equivalent to TA. ....

S. Nepal and M. V. Ramakrishna. Query processing issues in image (multimedia) databases. In Proc. 15th International Conference on Data Engineering (ICDE), pages 22--29, March 1999.

....1999) calls it a scoring rule over the tuple of id, score i pairs obtained from each of the n features. Thanks to the wide body of recent research on rules and formal techniques to express and compute this combination function (Fagin and Wimmers, 1999; Fagin, 1999; Adali et al. 1998; Nepal and Ramakrishna, 1999) our hypothetical system will have a rich collection of ways to use aggregate ranking functions. Thus, given the lack of access to the value or the structure of a feature, the system treats individual features as a black box with very poor support. However, it provides a wide variety of ....

Nepal, S. and Ramakrishna, M. (1999). Query processing issues in image (multimedia) databases. In Proc. 15th International Conference on Data Engineering, pages 22--29.

No context found.

S. Nepal and M. V. Ramakrishna. Query processing issues in image (multimedia) databases. In ICDE 1999.

No context found.

S. Nepal and M. V. Ramakrishna. Query processing issues in image (multimedia) databases. In ICDE 1999.

No context found.

Nepal, S., Ramakrishna, M.V.: Query processing issues in image (multimedia) databases. In: ICDE. (1999) 22--29

No context found.

Surya Nepal and M. V. Ramakrishna. Query processing issues in image (multimedia) databases. In ICDE, pages 22--29. IEEE Computer Society, 1999.

No context found.

Nepal, S., Ramakrishna, M.V.: Query processing issues in image (multimedia) databases. In: ICDE. (1999) 22--29

No context found.

S. Nepal and M. V. Ramakrishna. Query processing issues in image(multimedia) databases. In ICDE, pages 22--29, 1999.

No context found.

S. Nepal and M. V. Ramakrishna. Query processing issues in image(multimedia) databases. In ICDE, pages 22--29, 1999.

No context found.

Nepal S, Ramakrishna M (1999) Query processing issues in image (multimedia) databases. In: Proceedings of the 15th international conference on data engineering, Sydney, Australia, 23--26 March 1999, pp 22--29

No context found.

Nepal, S., Ramakrishna, M. Query processing issues in image (multimedia) databases. In ICDE'99 Proc. of the 15 International Conference on Data Engineering. March 23-26. Sydney, Australia. p. 2229. IEEE Computer Society, 1999.

No context found.

S. Nepal and M. V. Ramakrishna. Query processing issues in image (multimedia) databases. In Proc. of the 1999.

No context found.

Surya Nepal and M. V. Ramakrishna. Query processing issues in image (multimedia) databases. In ICDE, Sydney, Austrialia, 1999.

No context found.

Surya Nepal and M. V. Ramakrishna. Query processing issues in image (multimedia) databases. In ICDE, Sydney, Austrialia, 1999.

No context found.

S. Nepal and M. V. Ramakrishna. Query processing issues in image (multimedia) databases. In Proceedings of the IEEE International Conference on Data Engineering, 1999.

First 50 documents

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC