H. Lu, B. Ooi, and K. Tan. On Spatially Partitioned Temporal Join. In Proceedings of the 20th VLDB Conference, pages 546--557, 1994.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....is joined, the records that may possibly join with some records in the unprocessed partitions are retained in the join buffer. 20] used this dynamic partitioning algorithm while utilizing the Time Index [7] to determine the exact partitioning intervals so that each partition fits in memory. [16] proposed a (plain) T Join algorithm based on spatial partitioning. Here a record s interval i is mapped to a point (i:start, i:end i:start) in a two dimensional space. These points are then indexed by an R tree like method (the TP Index [21] which partitions the space. However, a partition in ....

....The plane sweep algorithm based on MVBT is only implemented for the GT Join, while the pipelined sortmerge join and the B tree join are only implemented for the GE Join. The reasons are discussed in sections 4 and 5. We also implemented a spatially partitioned GT Join using the approach in [16]. For each implemented algorithm, we mention the section where the algorithm is discussed. 6.2 Experimental Setup The algorithms were implemented in C and C using GNU compilers. The programs were run on a Sun Enterprise 250 Server machine with two UltraSPARC II processors using Solaris 2.8. To ....

[Article contains additional citation context not shown here]

H. Lu, B. Ooi and K. Tan, "On Spatially Partitioned Temporal Join", Proc. of VLDB, 1994.

....algorithm take the number of long lived tuples into consideration, which renders its performance insensitive to the number of long lived tuples. However, it relies on a pre existing temporal histogram. Lu, Ooi, and Tan described another range partitioning algorithm for computing temporal joins [LOT94] This algorithm is applicable to Theta joins, where a result tuple is produced for each pair of input tuples with overlapping valid time intervals. Their approach is to map intervals to a two dimensional plane, which is then partitioned into regions. The join result is produced by computing the ....

....joins, again investigating the effectiveness of both explicit attribute indexing and timestamp indexing. While a large number of timestamp indexes have been proposed in the literature [ST99] and there has been some work on temporal joins that utilize temporal or spatial indexes [EWK90, SE96, LOT94, ZTS02] a comprehensive empirical comparison is needed of these algorithms. Orthogonally, more sophisticated techniques for temporal database implementation should be considered. In particular, we expect specialized temporal database architectures to have a significant impact on query ....

H. Lu, B.-C. Ooi, and K.-L. Tan. On Spatially Partitioned Temporal Join. In Proceedings of the Conference on Very Large Databases, pages 546--557, Santiago, Chile, 1994.

....joined, the records that may possibly join with some records in the unprocessed partitions are retained in the join buffer. SE96] uses this dynamic partitioning algorithm while utilizing the Time Index ( EWK90] to determine the exact partitioning intervals so that each partition fits in memory. [LOT94] proposes a T Join algorithm based on spatial partitioning. Here a record s interval i is mapped to a point (i:start, i:end i:start) in a two dimensional space. These points are then indexed by an R tree like method (the Time Polygon Index) which partitions the space. However, a partition in one ....

H. Lu, B. Ooi and K. Tan, "On Spatially Partitioned Temporal Join", Proc. of VLDB, pp. 546-557, 1994.

....joined, the records that may possibly join with some records in the unprocessed partitions are retained in the join buffer. SE96] uses this dynamic partitioning algorithm while utilizing the Time Index ( EWK90] to determine the exact partitioning intervals so that each partition fits in memory. [LOT94] proposes a T Join algorithm based on spatial partitioning. Here a record s interval i is mapped to a point (i:start, i:end i:start) in a two dimensional space. These points are then indexed by an R tree like method (the Time Polygon Index) which partitions the space. However, a partition in one ....

H. Lu, B. Ooi and K. Tan, "On Spatially Partitioned Temporal Join", Proc. of VLDB, pp. 546-557, 1994.

....For example, in Table 1, the surrogate of the Emp Sal relation is Employee and Sex is the nontemporal attribute. Salary is the temporal attribute, and Vs, Ve, Ts, and Te are the time attributes of the relation. In this work, we assume that the temporal domain is a sequence of discrete time instants[6]. Table 1: A temporal relation Emp Sal Employee Sex Salary Vs Ve Ts Te Mary F 50 0 12 0 3 John M 40 0 6 0 3 Peter M 60 0 5 0 2 Mike M 50 6 12 0 2 Peter M 60 3 6 3 now Mike M 60 6 12 3 now Mary F 50 0 3 4 5 Mary F 70 4 8 6 now John M 50 7 12 4 now 2.2 Canonical temporal operators In temporal ....

H. Lu, B. Ooi, and K. Tan. On Spatially Partitioned Temporal Join. In Proceedings of the 20th VLDB Conference, pages 546--557, 1994.

.... power of a database while query optimization is a critical issue in practicality of a temporal database management system(TDBMS) So far a number of strategies to process temporal operators for temporal join, which is one of the most expensive operations in TDBMS, were proposed[GS91, SSJ94, LOT94] In general, join evaluation algorithms fall into three basic categories, nested loops, sort merge, and partition based. Among them a partition based approach, which has a great potential for effective temporal join handling, is our main concern. However, without some replications, ....

....a mechanism which can identify the counterparts of a partition the partitions of a joining relation which need to be compared with a partition of the other joining relation is essential to a partition based temporal join method. We observe that the two dimensional space proposed in [LOT94] needs to be extended to three dimensional space to cover various types of temporal join. In this paper, we propose a three dimensional spatially partitioned temporal operation model (3DTOM) which can be applied to process a variety of temporal operations. Since join operation is generally ....

[Article contains additional citation context not shown here]

H. Lu, B. Ooi, and K. Tan. On Spatially Partitioned Temporal Join. In Proceedings of the 20th VLDB Conference, 1994.

.... the partition based algorithms pose interesting challenge. Under a partition based algorithm, temporal data are split into partitions. During the join process, a partition in one relation only needs to join with some, but not all, partitions of the other relation. However, as pointed out in [6], the performance bottleneck of partition based temporal join method is that those partitions containing so called long lived records (records whose time intervals span over a long period) need to be compared with almost all the other partitions. As a result, the savings obtained from the ....

....span over a long period) need to be compared with almost all the other partitions. As a result, the savings obtained from the reduction of comparisons among partitions may not be able to offset the overhead of partition based methods the cost of partitioning. To solve the problem, Lu, et al. [6] avoided the partitioning phase by clustering records into partitions based on the time attributes. To facilitate direct access to the partitions, a spatial index is built on the set of partitions. In this paper, we address a subproblem in designing partition based temporal joins the ....

[Article contains additional citation context not shown here]

H. Lu, B. Ooi, and K. Tan. On spatially partitioned temporal joins. In Proceedings of the 20th International Conference on Very Large Data Bases, pages 546--557, Santiago, Chile, August 1994.

No context found.

H. Lu, B. Ooi, and K. Tan. On Spatially Partitioned Temporal Join. In Proceedings of the 20th VLDB Conference, pages 546--557, 1994.

No context found.

H. Lu, B. Ooi and K. Tan, \On Spatially Partitioned Temporal Join", Proc. of VLDB, 1994.

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