S. Ramaswamy and P. C. Kanellakis. OODB indexing by class-division. In Proc. ACM SIGMOD Internat. Conf. on Management of Data, pages 139--150, 1995.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....an algorithm in terms of the number of I Os it performs and the number of disk block it uses. The dynamic interval management problem is the problem of maintaining a set of intervals such that, given a query interval I q , all intervals intersecting I q can be reported eciently. As discussed in [28, 29, 37], the problem is crucial for indexing constraints in constraint databases and in temporal databases. The key component of dynamic interval management is the ability to answer stabbing queries [29] Given a set of intervals, a stabbing query with a point q asks for all intervals containing q. By ....

S. Ramaswamy and P. Kanellakis. OODB indexing by class division. In A.P.I.C. Series, Academic Press, New York, 1995.

....We attribute the renewed interest in the fundamental results of [9] where it was shown that indexing in new database paradigms, such as constraint databases, and databases with class hierarchies, can be reduced to special cases of multidimensional range searching. In subsequent publications, [14, 13, 16, 17] there are presented asymptotically efficient dynamic algorithms for 2sided, 3 sided and interval management queries. An optimal solution for the interval management problem has recently been found [1] We also mention the work of [12] who use cost metrics similar to ours, to characterize the ....

S. Ramaswamy and P. C. Kanellakis. OODB indexing by class-division. SIGMOD Record (ACM Special Interest Group on Management of Data), 24(2):139--150, June 1995.

....it happens, by the golden ratio OE = Gamma1 p 5) 2 then any indexing scheme for range queries must have storage redundancy logarithmic in the number of points if it is to achieve access overhead less than the block size. Before [6] there was substantial work on the indexing problem [9, 14, 15, 18, 21]. Most of this work involves upper bounds, and is therefore mainly concerned with the analysis of the searching aspect of the problem. There is however a (0; 0) 1; OE) b) a) GammaOE; 1) Figure 1: The Fibonacci lattice and F8 . notable result on lower bounds that appears in the last ....

S. Ramaswamy and P. C. Kanellakis. OODB Indexing by Class Division. In Proc. 12th ACM SIGACT-SIGMOD-SIGART Symposium on Principles of Database Systems, pp:233--243, 1993.

....this kind of hierarchy to domain mapping yields a significant query performance enhancement in a multikey type index. 1 Introduction Multikey index structures have been discussed as an alternative to traditional, B tree based search structures in the context of type hierarchy indexing (see [RK95], KRVV93] KM94] and [MP95] The key idea is to incorporate the type hierarchy structure of a given database scheme into a standard multikey index implementation in such a way that the hierarchy is mapped to one of the multikey index domains (called type domain in the sequel) The result is a ....

.... Related work on type hierarchy indexing includes two proposals based on standard B tree technology: the straightforward solution of maintaining one B tree per indexed type (called single class index in [KKD89] and an approach on replication of OIDs (called Class Division in [RK95]) Other proposals extend B trees, e.g. the Class Hierarchy Index [KKD89] maintaining a common B tree for all types and an additional leave node organization scheme. Single type B trees nested according to the inheritance hierarchy are introduced in [LOL92] as H trees. CG trees ....

[Article contains additional citation context not shown here]

Sridhar Ramaswamy and Paris C. Kanellakis. OODB indexing by class-division. In Michael Carey and Donovan Schneider, editors, Proceedings of the 1995 ACM SIGMOD International Conference on Management of Data, volume 24 of SIGMOD Record, pages 139--150, San Jose, CA, June 1995. ACM Press.

.... are well suited for estimating predicate selectivities, and for estimating the number of tuples one relation (or class) produces when joined with another relation (or class) Object oriented statistics can be somewhat more complicated if the class hierarchy is taken into account, e.g. CCY94, RK95, SS94, XH94] However, these statistics are not well suited for long sequences of joins as embodied in path expressions. A cost based optimizer for path expressions may, for example, need to accurately estimate the number of Book.Author.Address.City paths in the database. In Lore we set a ....

S. Ramaswamy and P. Kanellakis. OODB indexing by class-division. In Proceedings of the ACM SIGMOD International Conference on Management of Data, pages 139--150, San Jose, California, May 1995.

....predicates can be found in [20, 6, 13, 14, 24, 10, 11, 8] b) For inheritance hierarchy: The access scope of a query against a class may include instances of the class and those of its subclasses. An index structure can support both instances in the same search index. Previous work includes [17, 1, 18, 19, 16, 22, 21], 2. Behavioral For OODB, queries may contain method invocation. The behavioral indexing technique is based on pre computation of method results and storing them into an index. Previous work on behavioral indexing includes [15, 12, 13, 14, 2, 7] The effects of two indexes could be entangled, that ....

Sridhar Ramaswamy and Paris C. Kanellakis. Oodb indexing by class-division. In Proceedings ACM-SIGMOD Conference on the Management of Data, pages 139--150, 1995.

....the query is the set of points in the data structure that are contained in ae. Range searching is a fundamental primitive in several large scale applications, including spatial databases and geographic information systems (GIS) 1, 4, 6, 9, 11] graphics [3] indexing in object oriented databases [5, 7], and constraint logic programming [7] When the data are too large to fit in main memory and must reside on disk, the Input Output (I O) communication can become a very severe bottleneck. In d dimensional space, we define a (s 1 ; s 2 ; s d ) sided range query, where each s i 2 f1; 2g, ....

....structure that are contained in ae. Range searching is a fundamental primitive in several large scale applications, including spatial databases and geographic information systems (GIS) 1, 4, 6, 9, 11] graphics [3] indexing in object oriented databases [5, 7] and constraint logic programming [7]. When the data are too large to fit in main memory and must reside on disk, the Input Output (I O) communication can become a very severe bottleneck. In d dimensional space, we define a (s 1 ; s 2 ; s d ) sided range query, where each s i 2 f1; 2g, to be an orthogonal range query with s ....

S. Ramaswamy and P. C. Kanellakis. OODB indexing by class division. In Proc. 1995 ACM International Conference on Management of Data, 1995.

....the design of specific indexing techniques. Indexing techniques so far proposed can be classified into three groups: inheritance indexing techniques, aggregation indexing techniques, integrated techniques [8] The first group includes techniques supporting queries along inheritance hierarchies [10, 11, 13, 14]. Most of those techniques have been defined as extensions of the class hierarchy index (CHindex) 11] The CH index is based on B trees and maintains a single index for all classes in y Recommended by Patrick O Neil 66 Elisa Bertino et al. the given inheritance hierarchy. The CH index ....

....Conversely, the H index [13] efficiently processes range predicates but reads more pages than CH index for single key predicates. As a result, a method combining the advantages of both CH and H trees has been proposed in [15] A different approach has been proposed by Kanellakis, Ramaswamy et al. [14] based on external dynamic 2 dimensional range searching techniques. The second group of techniques deals with the efficient evaluation of nested predicates. Most of those techniques are based on the same principle of the join index technique proposed by Valduriez in [17] In particular, they are ....

S. Ramaswamy and P.C. Kanellakis. OODB Indexing by Class-Division. In Proc. of the ACM SIGMOD Int. Conf. on Management of Data, pages 139--150, San Jose, California, 1995.

....the query is the set of points in the data structure that are contained in ae. Range searching is a fundamental primitive in several large scale applications, including spatial databases and geographic information systems (GIS) 9, 6, 1, 11, 4] graphics [3] indexing in objectoriented databases [5, 7], and constraint logic programming [7] When the data are too large to fit in main memory and must reside on disk, the Input Output (I O) communication can become a very severe bottleneck. In d dimensional space, we define a (s 1 ; s 2 ; s d ) sided range query, where each s i 2 f1; 2g, to ....

....structure that are contained in ae. Range searching is a fundamental primitive in several large scale applications, including spatial databases and geographic information systems (GIS) 9, 6, 1, 11, 4] graphics [3] indexing in objectoriented databases [5, 7] and constraint logic programming [7]. When the data are too large to fit in main memory and must reside on disk, the Input Output (I O) communication can become a very severe bottleneck. In d dimensional space, we define a (s 1 ; s 2 ; s d ) sided range query, where each s i 2 f1; 2g, to be an orthogonal range query with s i ....

S. Ramaswamy and P. C. Kanellakis. OODB indexing by class division. In Proc. ACM International Conference on Management of Data, 1995.

....as a simple path index. A preliminary version of Lore s query language, Lorel, was introduced in [QRS 95] Details of the current version of Lorel appear in [AQM 97] Needless to say, there has been a significant amount of work in indexing for object oriented databases, e.g. KKD89, SS94, RK95, KM92, CCY94, BG92, XH94] All of this work depends on the database having a fixed schema based on a known, strongly typed class hierarchy. In our environment we must take a different approach to indexing, since we do not have a fixed schema, and comparable objects may take on different types. ....

S. Ramaswamy and P. C. Kanellakis. OODB indexing by class-division. SIGMOD Record, 24(2):139--150, June 1995.

....The creation of the optimal set of indexes for the given query mix and database population is task of the system administrator as mentioned above. 5.1. 3 Efficient Searching in Hierarchies Several ways to index objects that are organized in a hierarchy have been discussed in the literature [8,10]. Most of these proposals focus on a small and static hierarchy where a user asks hierarchically restricted attribute queries. The most common technique is to map the position of an object in the hierarchy into a number such that all nodes in a sub tree are assigned a number from a known disjoint ....

S. Ramaswamy and P. Kanellakis. OODB Indexing by Class-Division. SIGMOD Conference 1995, 139-150.

....report some benchmarks of TPIE on sorting and matrix multiplication. This work, however, is mainly on providing a programming environment and not on performance comparisons between external memory algorithms and conventional algorithms. Also worth noting is the work by Ramaswamy and Kanellakis [30], who study the problem of indexing a class hierarchy in Object Oriented Databases. They propose a technique called class division, and show by experiments that in the average case, class division performs far less I O s than the class hierarchy index technique most popular today. Their ....

S. Ramaswamy and P. C. Kanellakis. OODB indexing by class-division. In Proc. ACM SIGMOD Internat. Conf. on Management of Data, pages 139--150, 1995.

....schemes. Although our framework for studying indexability is new, there are some previous results that fit into it rather naturally. Analyses along the lines we are suggesting here have been emerging in the past few years, most notably in the work of the late Paris Kanellakis and his collaborators [15, 23, 25, 27, 30]. Most of this work involves upper bounds, and is therefore mainly concerned with the analysis of the searching aspect of the problem. There are two exceptions. First, in a recent version of [15] there is an argument (proof of Lemma 2.7) that anticipates our Theorem 1, namely, that the access ....

S. Ramaswamy, P. C. Kanellakis. OODB Indexing by Class Division. In Proc. 12th ACM SIGACTSIGMOD -SIGART Symposium on Principles of Database Systems, pages 233--243, 1993.

....reported some benchmarks of TPIE on sorting and matrix multiplication. This work, however, is mainly on providing a programming environment and not on performance comparisons between external memory algorithms and conventional algorithms. Also worth noting is the work by Ramaswamy and Kanellakis [97], who study the problem of indexing a class hierarchy in Object Oriented Databases. Based on the insight of Kanellakis et al. 65] that the problem is a special form of two dimensional dynamic range searching in external memory, they propose a technique called class division, and show by ....

S. Ramaswamy and P. C. Kanellakis. OODB indexing by class-division. In Proc. ACM SIGMOD Internat. Conf. on Management of Data, pages 139--150, 1995.

....on one attribute. This index structure has optimal worst case storage and query performance, and optimal amortized insert time (i.e. averaged over a sequence of inserts) with performance very close to that of the B tree. By studying the techniques that he had used for indexing constraints [RK95], Paris was also able to develop index methods for object oriented databases, and at the time of his death he was investigating their application to temporal databases. 4 Fault Tolerant Parallel Computation Paris Kanellakis, among other researchers, sought to bridge the gap between abstract ....

S. Ramaswamy and P. C. Kanellakis. OODB Indexing by Class-Division. in 1995 ACM Symposium on the Management of Data, pp. 139--150. Fault-Tolerant Parallel Computation

No context found.

S. Ramaswamy and P. C. Kanellakis. OODB indexing by class-division. In Proc. ACM SIGMOD Internat. Conf. on Management of Data, pages 139--150, 1995.

No context found.

S. Ramaswamy and P. Kanellakis, OODB indexing by class division, A.P.I.C. Series, Academic Press, New York, 1995.

No context found.

S. Ramaswamy and P. C. Kanellakis. OODB Indexing by Class Division. In Proc. 12th ACM SIGACT-SIGMOD-SIGART Symposium on Principles of Database Systems, pages 233--243, 1993.

No context found.

S. Ramaswamy and P. Kanellakis, OODB indexing by class division, A.P.I.C. Series, Academic Press, New York, 1995.

No context found.

S. Ramaswamy and P. Kanellakis, OODB indexing by class division, A.P.I.C. Series, Academic Press, New York, 1995.

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