Slot index spatial join

Abstract

Abstract-Efficient processing of spatial joins is very important due to their high cost and frequent application in spatial databases and other areas involving multidimensional data. This paper proposes slot index spatial join (SISJ), an algorithm that joins a nonindexed data set with one indexed by an R-tree. We explore two optimization techniques that reduce the space requirements and the computational cost of SISJ and we compare it, analytically and experimentally, with other spatial join methods for two cases: 1) when the nonindexed input is read from disk and 2) when it is an intermediate result of a preceding database operator in a complex query plan. The importance of buffer splitting between consecutive join operators is also demonstrated through a two-join case study and a method that estimates the optimal splitting is proposed. Our evaluation shows that SISJ outperforms alternative methods in most cases and is suitable for limited memory conditions.

TRANSFORMERS: Robust spatial joins on non-uniform data distributions

“…As I B is built based on I A , the bounding boxes are aligned leading to less overlap and the synchronous join therefore has to compare fewer bounding boxes. Extensions to the basic approach use sampling to build the R-Trees faster [18] or avoid memory thrashing [19]. Sampling the spatial datasets is also used to make the spatial join interactive [20].…”

An effective encoding scheme for spatial RDF data

“…First, the Rtree join algorithm [9] can be used in the case where both spatially joined variables involved in the Filter condition are instantiated directly from the base data and do not come as outputs of other query operators. Second, we use the SISJ algorithm [19] for the case where the R-tree can be used only for one variable. Finally, we implemented a spatial hash join (SHJ) algorithm [18] for the case where both inputs of the spatial join filter condition are output by other operators.…”

“…Algorithm(s) to Consider L and R sorted on entity IDs SMJ (Section 6.3) L and R results of (?si hasGeometry ?gi) SMJ or R-tree Join [9] L sorted on entity IDs SMJ (Section 6.3), SISJ [19], R result of a pattern (?s2 hasGeometry ?g2) or Index Nested Loops L unsorted SHJ-ID (Section 6.4), SISJ [19] R result of a pattern (?s2 hasGeometry ?g2) or Index Nested Loops L and R unsorted SHJ-ID, SHJ [18] or Nested Loops Depending on whether the inputs of the join are indexed, sorted, or unsorted, there are different algorithms to be considered. If both join inputs come ordered by the IDs of the spatial entities to be joined, then SMJ (Section 6.3) is the algorithm of choice.…”

“…In the special case where both inputs are the results of ?si hasGeometry ?gi patterns applied on the entire set of triples, besides of applying SMJ on the SPO (or SOP) index, we can apply an R-tree self-join [9] on the R-tree index (see Section 4). When just one of the inputs, e.g., R, is a result of a ?si hasGeometry ?gi pattern, besides SMJ, we can also apply the SISJ algorithm [19] (see Section 4). In this case, we also consider Index Nested Loops join using the R-tree, by applying one spatial range query for each tuple of the other input, e.g., L. This is expected to be cheap only when L is very small.…”

“…For estimating the selectivity of spatial query components, we use grid-based statistics, similar to previous work on spatial query optimization (e.g., see [19]). Specifically, we take advantage of statistics that are obtained by the spatial encoding phase of the entity IDs.…”