T. B. Pedersen and C. S. Jensen. Multidimensional Data Modeling for Complex Data. In Proceedings of the Fifteenth International Conference on Data Engineering, 1999.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....dimensions [9] and concluded that state oriented warehouses allow easier analytical processing and even better query performance than observed in regular events warehouses. Our formal approach to managing temporal consistency (described in section 4) is state oriented, too. Pedersen and Jensen [11] describe features that entire DWH data models should have (including a requirement to handle changes in data over time) and evaluate previously proposed models. In the discipline of temporal data warehouses a lot of research was done in the context of temporal view maintenance, e.g. 13] An ....

....(e.g. an old monthly sales total) even if the new information indicates that the old sales total was incorrect. Traditional DWHs ignoring the revelation time of late arriving records will invalidate cash flows and statistics. The proposed model enables explicit hierarchies in the time dimensions [11] to aid the user in navigation. It allows multiple hierarchies in the time dimension based on the load timestamp, e.g. days could roll up into weeks or months. To our knowledge it is the first DWH model that handles the change in data over time systematically by adding validity periods and ....

T.B. Pedersen and C.S. Jensen. Multidimensional Data Modeling for Complex Data. In Proc. of 15 ICDE, IEEE Computer Society, pp. 336-345, Sydney, Australia, 1999.

....required every pair of elements of a given category to have ancestors in the same set of categories, a restriction referred to as homogeneity. For example, in a homogeneous dimension we cannot have some cities that rollup to provinces and some to states. A number of researchers and practitioners [11, 8, 13, 6] have dropped this restriction over the past few years, yielding heterogeneous dimensions, which are needed to represent more naturally and cleanly many practical situations. Moreover, heterogeneous dimensions permit more e#cient storage of data by having fewer categories. A smaller number of ....

....must rewrite a cube view as another query that refers to pre computed cube views. The process of finding such rewritings is known in the OLAP world as aggregate navigation [9] The notion of summarizability was introduced to study aggregate navigation in statistical objects and OLAP dimensions [12, 11, 13, 6]. As originally stated, summarizability refers to whether a simple aggregate query (usually called summarization or consolidation) correctly computes a single category cube view from another precomputed singlecategory cube view, in a particular database instance. In previous work [6] we extended ....

[Article contains additional citation context not shown here]

T. B. Pedersen and C. S. Jensen. Multidimensional data modeling for complex data. In Proceedings of the 15th IEEE International Conference on Data Engineering, Sydney, Australia, 1999.

....hierarchy of conceptual clusters implicit in the original table. To handle those conceptual clusters in a multidimensional way and, furthermore, query them, we need to introduce a proper multidimensional model. Several researchers have proposed variations of a multidimensional model [Fir98, IH94, PJ99, Vas98, CT97, AGS97] and our work will center towards an extension of it so as to include clusters and concepts. From our description of hierarchical clustering it seems reasonable for our task to use an agglomerative algorithm that incorporates a hierarchy, with clusters organized in a cluster ....

Torben Bach Pedersen and Christian S. Jensen. Multidimensional Data Modeling for Complex Data. In Proc. of the 15th Int'l Conference on Data Engineering, (ICDE), pages 336--345, 23--26 March 1999.

....claimed to be as powerful as Relational algebra. Multidimensional operations like roll up can be built by composition of them. 8] uses Description Logics to describe semantics of multidimensional operators. Here, a cube is defined as an object which is associated to cells of similar form. [12] provides a formalism and an algebra that is closed and, at least, as strong as Relational algebra with aggregation functions. Finally, 18] also presents a complete and sound algebra. With regard to relationships among data in multidimensional schemas, 9] states that the fact table has a ....

T. B. Pedersen and C. S. Jensen. Multidimensional data modeling for complex data. In Proc. of 15th Int. Conf. on Data Engineering (ICDE), pages 336--345. IEEE Computer Society, 1999.

.... [1, 3, 13, 14, 17] Some models provide statistical objects where a structured hierarchy is related to an explicit aggregation function on a single measure supporting a set of queries [20] To model dimensions of complex structures, several models were made in an object oriented framework [3, 16, 21]. Also, some proposals exploits the temporal nature of the multidimensional modelling [10, 15, 16] Most of these proposals introduce constraints and specific modelling choices as ROLAP, MOLAP and OOLAP. Nevertheless, in [22] the authors provide a full 79 conceptual approach through the starER ....

.... is related to an explicit aggregation function on a single measure supporting a set of queries [20] To model dimensions of complex structures, several models were made in an object oriented framework [3, 16, 21] Also, some proposals exploits the temporal nature of the multidimensional modelling [10, 15, 16]. Most of these proposals introduce constraints and specific modelling choices as ROLAP, MOLAP and OOLAP. Nevertheless, in [22] the authors provide a full 79 conceptual approach through the starER model, which combines the star structure with the semantically rich constructs of the ER model. In ....

Pedersen T.B., Jensen C.S., "Multidimensional Data Modeling for Complex Data", ICDE'99, San Diego (California, USA), March 1999.

.... measures and dimension members Complex measures ffl Support of structured measures ffl Support of derived measures ffl Additivity of measures Query formalism ffl Type of formalism (i.e. algebra or calculus) ffl Ad hoc hierarchies ffl User defined aggregates [Ped00] see [PJ98] and [PJ99] for preliminary work) presents eleven requirements (found in clinical data warehousing) for multidimensional data models, and evaluates twelve preexisting data models against them. Those presented in [AGS97] Dyr96] GBLP96] Kim96] LW96] GL97] CT98b] DT97] Leh98] and [Vas98] are ....

....at LL. If we look at Facts Cells (at IL) which are not explicitly defined, we see that they are a tuple of Measures identified by the bottom levels of the different dimension lattices. Pedersen (Extended Multidimensional Data Model) Besides a classification of multidimensional models, PJ98] [PJ99], and [Ped00] already referenced in section 2, also present an Extended Multidimensional Data Model (EMDM) is presented. After the definition of the requirements (most of them refer to semantics) for the usage of a multidimensional model in a clinical context, and the verification that none of ....

T. B. Pedersen and C. S. Jensen. Multidimensional data modeling for complex data. In Proc. of 15th Int. Conf. on Data Engineering (ICDE), pages 336--345. IEEE Computer Society, 1999.

....analysis dimensions. Conclusions are in section 5, followed by acknowledgements and bibliography. 2 Semantic problems in present multidimensional modeling This section outlines some problems found in existing multidimensional models. Some of them were already identified in [SR91] Leh98] and [PJ99]. Even though [SR91] can be found out of place, most of the problems it identifies in statistical modeling are also applicable in multidimensional context. The problems, related to modeling dimensions, are grouped into five sections. 2.1 Aggregation levels graph At first glance, one could think ....

....must be connected and show parent child relationships between attributes. LAW98] imposes the existence of a common top aggregation level (called All) defining a lattice of aggregation levels for every analysis dimension; and identifies relationships between levels as functional dependencies. [PJ99] also identifies multiple aggregation paths in the same dimension, and presents the different aggregation levels forming a lattice, being related by greater than relationships (meaning logical containment of the elements at one level into those at the other) It could also be the case that our ....

[Article contains additional citation context not shown here]

T. B. Pedersen and C. S. Jensen. Multidimensional data modeling for complex data. In Proc. of 15th Int. Conf. on Data Engineering (ICDE), pages 336--345. IEEE Computer Society, 1999.

....the rest of the paper as follows. Next a review of the related work is given in Section 2. After that we present the basis of hierarchy types in Section 3. In Section 4 the new dependencies are studied. Finally, the conclusions are presented in Section 5. 2 Related Work Pedersen and Jensen [PeJe99] discuss requirements for multidimensional data models. Five of their nine requirements are for dimension hierarchies: 1) explicit hierarchies in dimensions, 2) support for multiple hierarchies in a dimension, 3) correct aggregation of data, 4) support for non strict hierarchies, and 5) data ....

Pedersen, T. and Jensen, C.: Multidimensional data modeling for complex data, Proceedings of the International Conference on Data Engineering (ICDE'99), 1999.

....and we define a conceptual multidimensional data model. We point out that in contrast to the relational data model there is no standard multidimensional data model yet. In fact, there is a variety of competing proposals; recent overviews and comparisons of multidimensional models appear in [BSHD98, PJ99, SBH99, VS99]. In this paper we take up and formalize the multidimensional terminology used in [HLV00] A data warehouse can be understood as a multidimensional database, whose atomic information units are given by facts. A fact can be perceived as a point in a multidimensional space to which some measurable ....

....capital. Thus, the computation of measure ROI from turnover, costs, and invested capital involves sum, difference, and division. We note that, in general, the derivation order of a measure schema may not induce a hierarchy and not even a partial order. 0 As already argued, for example, in [GMR98, LS97, PJ99], not all possible aggregations of measures within a certain application scenario make sense in general. For example, given a group of customers at a specific date, summing over as well as taking averages of account balances may be perfectly reasonable; however, summing up account balances over ....

T.B. Pedersen, Ch.S. Jensen, "Multidimensional Data Modeling for Complex Data," Proc. ICDE 1999, 336--345.

....OLTP data model into a dimensional model) Note that we describe multidimensional data on a conceptual level, which allows us to translate the model into multidimensional arrays as well as into the relational data model. In the recent years, several multidimensional data models have been proposed [37, 25, 5, 27, 21, 15, 4, 32]. We shortly review the basic principles of multidimensional data and discuss some outstanding aspects and differences between these approaches. A more in depth comparison is provided by Pedersen and Jensen in [32] 5 The basic unit of interest in a data warehouse are measures or facts (e.g. ....

....several multidimensional data models have been proposed [37, 25, 5, 27, 21, 15, 4, 32] We shortly review the basic principles of multidimensional data and discuss some outstanding aspects and differences between these approaches. A more in depth comparison is provided by Pedersen and Jensen in [32]. 5 The basic unit of interest in a data warehouse are measures or facts (e.g. sales) which represent countable, semi summable, or summable information [26] concerning a business process. An instance of a measure is called measure value. A measure can be analyzed from different perspectives, ....

[Article contains additional citation context not shown here]

T. B. Pedersen and C. S. Jensen. Multidimensional Data Modeling for Complex Data. In ICDE'99, Proceedings of 15th International Conference on Data Engineering, March 23-26, 1999, Sydney, Australia, pages 336--345. IEEE Computer Society Press, 1999.

....engines to perform data mining, and using the distributed OLAP infrastructure to scale data mining. We have also shown the power of the enhanced multilevel and multidimensional association rules for ecommerce applications. OLAP technology has gained increasing popularity in data warehousing [1,2,5,8]. However, the issues regarding the use of OLAP servers as distributed computation engines have not been studied. There exist a number of previous efforts on association rule mining from databases or other data sets [9,10,11] Several of these efforts are based on cube structures using OLAP [6] ....

Torben Bach Pedersen, Christian S. Jensen, "Multidimensional Data Modeling for Complex Data", Proc. ICDE'99, 1999.

....using the starER model. The financial and the administration department are outside the customer dimension as this is dependent on design decisions and semantic issues. 5 Evaluation of the starER model In this section we evaluate the starER model based on a welldefined set of criteria [11] and later on, we compare the proposed model to other conceptual data warehouse models [5] 11] present a set of evaluation criteria for data warehouse models. We adopt them here to evaluate the starER model. The criteria allow us to decide whether or not a model is adequate for data warehouse ....

....customer dimension as this is dependent on design decisions and semantic issues. 5 Evaluation of the starER model In this section we evaluate the starER model based on a welldefined set of criteria [11] and later on, we compare the proposed model to other conceptual data warehouse models [5] [11] present a set of evaluation criteria for data warehouse models. We adopt them here to evaluate the starER model. The criteria allow us to decide whether or not a model is adequate for data warehouse modeling, in terms of correctness, modeling power and efficiency in information capturing, ....

[Article contains additional citation context not shown here]

Pedersen, T. B., and Jensen, C. S., 1998. Multidimensional Data Modeling of Complex Data. Proceedings of the 15 th IEEE International Conference on Data Engineering (ICDE 99), Sydney, Australia.

....design,andphysical schema design. Concerning data warehouse design, there is a general agreement that at least a conceptual or logical modeling activity should precede the actual implementation [WB97, AGS97, CT98, GMR98] Typically, the modeling activity is based on a multidimensional model (see [BSHD98, VS99, PJ99] for comparisons of various multidimensional models) whereas the implementation is carried out either within relational or multidimensional databases [CD97] However, most of these models were developed without an embedding into a design process and thus without guidelines on how to use them or ....

....branch if the customerType is #business customer# and to profession otherwise. Thus, customerType is a join level that satisfies our requirements. In general, it might be necessary to introduce such a join level explicitly. Definition of summarizability constraints As already argued in, e.g. [LS97, GMR98, PJ99], not all possible aggregations of measures within a certain application scenario make sense in general. For example, given a group of customers, summing over as well as taking averages of account balances may be perfectly reasonable; similarly, the computation of average ages may be reasonable, ....

[Article contains additional citation context not shown here]

T.B. Pedersen, Ch.S. Jensen, "Multidimensional Data Modeling for Complex Data," Proc. ICDE 1999, 336-- 345.

No context found.

T. B. Pedersen and C. S. Jensen. Multidimensional Data Modeling for Complex Data. In Proceedings of the Fifteenth International Conference on Data Engineering, 1999.

No context found.

T. B. Pedersen and C. S. Jensen. Multidimensional Data Modeling for Complex Data. TimeCenter TR-37, #www.cs.auc.dk/TimeCenter#, 1998.

No context found.

T. B. Pedersen and C. S. Jensen. Multidimensional Data Modeling for Complex Data. In Proceedings of the Fifteenth International Conference on Data Engineering,1999.

No context found.

T. B. Pedersen and C. S. Jensen. Multidimensional Data Modeling for Complex Data. In Proceedings of the Fifteenth International Conference on Data Engineering, pp. 336--345, 1999.

No context found.

T. B. Pedersenand C. S. Jensen. Multidimensional Data Modeling for Complex Data. In Proceedings of ICDE, pp. 336-- 345, 1999.

.... traditional types of DBMSs, including automatic application of the pre specified aggregation functions (automatic aggregation) Rafanelli et al. 1990, Thomsen, 1997] visual querying [Thomsen, 1997, Thomsen, 1999] and good query performance due to the use of preaggregation [Gupta et al. 1995, Pedersen et al. 1999b] Additionally, the dimensional approach is most often a natural fit for data analysis problems. To be able to capture the complex data found in many real world applications, the data model for the OLAP system must be able to handle irregular dimension hierarchies [Pedersen et al. 1999a, ....

....1995, Pedersen et al. 1999b] Additionally, the dimensional approach is most often a natural fit for data analysis problems. To be able to capture the complex data found in many real world applications, the data model for the OLAP system must be able to handle irregular dimension hierarchies [Pedersen et al. 1999a, Pedersen et al. 1999b] that do not fit the balanced tree hierarchies supported by current OLAP systems. More specifically, the data model must support non strict hierarchies where lowerlevel items may have several parents in a higher dimension level, and non covering hierarchies, where paths ....

[Article contains additional citation context not shown here]

T. B. Pedersen and C. S. Jensen. Multidimensional Data Modeling for Complex Data. In Proceedings of the Fifteenth International Conference on Data Engineering, pp. 336--345, 1999.

No context found.

T. B. Pedersen and C. S. Jensen. Multidimensional data modeling for complex data. In Proceedings of the 15th IEEE International Conference on Data Engineering, Sydney, Australia, 1999.

No context found.

T.B Pedersen and C. Jensen. Multidimensional data modeling for complex data. Proceedings of IEEE/ICDE'99, 1999.

No context found.

Pedersen, T.B., Jensen, C.S.: Multidimensional data modeling for complex data. 15th International Conference on Data Engineering (ICDE '99), Sydney, Australia, March 1999

No context found.

Pedersen T. B., Jensen C. S., "Multidimensional data modeling for complex data", In Proc. of 15 Int. Conf. on Data Engineering (ICDE), IEEE Computer Society, 1999.

No context found.

T. B. Pedersen and C. S. Jensen. Multidimensional data modeling for complex data. In Proc. of 15th Int. Conf. on Data Engineering (ICDE), pages 336--345. IEEE Computer Society, 1999.

No context found.

T.B. Pedersen and C.S. Jensen, "Multidimensional Data Modeling of Complex Data," Proc. 15th IEEE Int'l Conf. Data Eng. (ICDE 99), IEEE CS Press, Los Alamitos, Calif., 1999, pp. 336-345.

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