D. Barbara, W. DuMouchel, C. Faloutsos, P. Hass, J. M. Hellerstein, Y. Ioannidis, H. Jagadish, T. Johnson, R. Ng, V. Poosala, K. Ross, and K. Sevcik. The new jersey data reduction report. Bulletin of the Technical Committee on Data Engineering, 20(4):3--45, December 1997.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....forms of aggregate queries. Besides the type of queries supported, another crucial aspect of an approximate query processing technique is the employed data reduction mechanism; that is, the method used to obtain synopses of the data on which the approximate query execution engine can then operate [9]. The methods explored in this context include sampling and, more recently, histograms and wavelets. Sampling based techniques are based on the use of random samples as synopses for large data sets. Sample synopses can be either precomputed and incrementally maintained (e.g. 1, 51] or they ....

D. Barbara, W. DuMouchel, C. Faloutsos, P.J. Haas, J.M. Hellerstein, Y. Ioannidis, H.V. Jagadish, T. Johnson, R. Ng, V. Poosala, K.A. Ross, and K.C. Sevcik. "The New Jersey Data Reduction Report". IEEE Data Engineering Bulletin, 20(4):3--45, December 1997. (Special Issue on Data Reduction Techniques).

....high dimensional feature spaces (HDFS) it must be used in conjunction with a dimensionality reduction technique in order to exploit the correlations in data and hence achieve further scalability. This approach is commonly used in both multimedia retrieval ( 43, 103, 76, 142] and data mining ([47, 8, 49]) applications. The idea is to first reduce the dimensionality of the data and then index the reduced space using a multidimensional index structure [43] Most of the information in the dataset is condensed to a few dimensions (the first few principal components (PCs) by using principal component ....

D. Barbara, W. DuMouchel, C. Faloutsos, P. Haas, J. Hellerstein, Y. Ionnidis, H. Jagadish, T. Johnson, R. Ng, V. Poosala, K. Ross, and K. Sevcik. The new jersey data reduction report. Data Engineering, 20(4), 1997.

.... to keep all relevant tuples in memory (and frequent access to hard disk will be too slow when arrival rates are high) In order to deal with resource limitations in a graceful way, returning approximate query answers instead of exact answers has emerged as a promising approach to save resources [4]. In data stream processing systems, one way of approximating query answers is to shed load, for example, by dropping tuples before they naturally expire (i.e. leave the window) or even before they reach the operator. The current state of the art consists of two main approaches. The first relies ....

D. Barbara, W. DuMouchel, C. Faloutsos, P. J. Haas, J. M. Hellerstein, Y. E. Ioannidis, H. V. Jagadish, T. Johnson, R. T. Ng, V. Poosala, K. A. Ross, and K. C. Sevcik. The New Jersey data reduction report. IEEE Data Engineering Bulletin, 20(4):3--45, 1997.

....clusters or outliers, others might not. To date, one has no way of knowing that, unless one resorts to executing the appropriate algorithms on the data. A powerful paradigm to improve the efficiency of a given data analysis task is to reduce the size of the dataset (often called Data Reduction [4]) while keeping the accuracy loss as small as possible. The difficulty lies in designing the appropriate approximation technique for the specific task and datasets. Sampling [48, 36] is a well recognized and widely used statistical technique. In the context of clustering in databases, both ....

D. Barbara, C. Faloutsos, J. Hellerstein, Y. Ioannidis, H. V. Jagadish, T. Johnson, R. Ng, V. Poosala, K. Ross, K.C. Sevcik. The New Jersey Data Reduction Report. In Data Engineering Bulletin, 9f1996

....e.g. the sum of sales. In contrast, our work achieves a reduction in data volume, while still allowing all dimensions to remain in the warehouse, by offering mechanisms for selectively aggregating data to higher levels. Previous work has also dealt with general techniques for data reduction [2] such as wavelets, sampling, and aggregation. These types of work differ in focus from ours, which offers rule based specifications for when data reduction, in our case using aggregation, should take place, allowing the reduction process to be automatic and transparent to the user. The aggregation ....

D. Barbara, W. DuMouchel, C. Faloutsos, P. J. Haas, J. M. Hellerstein, Y. Ioannidis, H. V. Jagadish, T. Johnson, R. Ng, V. Poosala, K. A. Ross, and K. C. Sevcik. The New Jersey Data Reduction Report. IEEE Data Engineering Bulletin, 20(4):3--42, 1997.

....using singular value decompositions of a term document matrix in information retrieval to find hidden structure. This has parallels in Karhunen Loeve expansions in signal representation and principal component analysis in statistics. A surview of such data reduction techniques appears in [Barbara et al. 1997]. Section 4 most directly deals with this perspective. 2 Motivating Domains As the size and complexity of datasets gathered from large scale simulations and high resolution observations increases, there is a significant push towards developing tools for interpreting these datasets. In spite of ....

Barbara, D., DuMouchel, W., Faloutsos, C., Haas, P., Hellerstein, J., Ioannidis, Y., Jagadish, H., Johnson, T., Ng, R., Poosala, V., Ross, K., and Sevcik, K. (December 1997). The New Jersey Data Reduction Report. Bulletin of the IEEE Technical Committee on Data Engineering, Vol. 20(4):pp. 3--45.

....one resorts to executing the appropriate algorithms on the data. A powerful paradigm to improve the efficiency of a given data analysis task is to reduce the size of the dataset (often called Data This research has been supported by NSF CAREER Award 9984729 and NSF IIS 9907477. Reduction [3]) while keeping the accuracy loss as small as possible. The difficulty lies in designing the appropriate approximation technique for the specific task and datasets. In this paper, we propose to use biased sampling as a data reduction technique to efficiently provide approximate solutions to data ....

D. Barbara, C. Faloutsos, J. Hellerstein, Y. Ioannidis, H. V. Jagadish, T. Johnson, R. Ng, V. Poosala, K. Ross, and K.C. Sevcik. The new jersey data reduction report. Data Engineering Bulletin, September 1996.

....In huge data warehouses it often makes sense to summarize a dataset in order to reduce its size. An approximate answer to a query computed using the summary may be feasible when an exact answer using the full dataset may be infeasible. A survey of such data reduction techniques is presented in [13]. Techniques for visualizing large multidimensional datasets are presented in [14] Our techniques enable one to visually identify, for any pair of dimensions, regions where the two dimensional distribution is not explainable as the independent combination of onedimensional distributions. 3 JAM ....

D. Barbara et al. The New Jersey data reduction report. Data Engineering Bulletin, 20(4):3--45, 1997.

....p j m[ j; a( p; j) 1] A bit string of length b = d Gamma1 j=0 b j represents each cell. Such a bit string is the concatenation of the bit strings of the interval numbers of the cell. Finally, the approximation of p is the bit string of the 4 0 1 2 3 0 1 2 3 m[1,0] m[1,1] m[1,2] m[1,3] m[1,4] m[0,4] 01 10 00 10 10 10 11 10 01 01 00 01 10 01 11 01 01 00 00 00 10 00 11 00 01 11 00 11 10 11 11 11 p q dist 2 uBnd 2 lBnd 2 cell(p) m[j,a(p,j) m[j,a(p,j) 1] a) b) Figure 1: Illustration of the VA File cell that contains p. Notice that for large d, the volume of a ....

....lower dimensionality that preserves the (ordering of) distances between vectors in the original space as well as possible. The objective, at least in theory, is that those tree based indexing techniques work , i.e. are better than linear. A typical dimensionality reduction technique out of many [3] is Singular Value Decomposition (SVD) Its results are relatively good. On the other hand, SVD is rather expensive, its complexity is O(N d 2 ) with N being the number of objects and d being the dimensionality. The ugly thing is that, after modifications, SVD must be run again for the entire ....

D. Barbara, W. DuMouchel, C. Faloutsos, P. J. Haas, J.M. Hellerstein, Y. Ioannidis, H.V. Jagadish, T. Johnson, R. Ng, V. Poosala, K.A. Ross, and K. C. Sevcik. The New Jersey data reduction report. Data Engineering, 20(4):3--45, 1997.

....query evaluation techniques described in Section 4. Section 5 defines different measures for the loss of result quality. We use these measures in Section 6 to evaluate the different techniques. Section 7 discusses related work, and Section 8 concludes. 0 1 2 3 0 1 2 3 m[1,0] m[1,1] m[1,2] m[1,3] m[1,4] m[0,1] 01 10 00 10 10 10 11 10 01 01 00 01 10 01 11 01 01 00 00 00 10 00 11 00 01 11 00 11 10 11 11 11 p q dist 2 uBnd 2 lBnd 2 cell(p) m[j,a(p,j) m[j,a(p,j) 1] a) b) Fig. 1. Illustration of the VA File 2 Preliminaries To ease presentation, we explicitly ....

....reduction. The principle is as follows: one tries to find a mapping of the vectors to vectors in a space with lower dimensionality that preserves the (ordering of) distances between vectors in the original space as well as possible. A typical dimensionality reduction technique out of many [2] is Singular Value Decomposition (SVD) Its results are relatively good, but it is rather expensive. 13] proposes an extension to SVD that operates on aggregates. Performance improves by an order of magnitude, and the loss of quality is acceptable. In this current context, dimensionality ....

D. Barbara, W. DuMouchel, C. Faloutsos, P. J. Haas, J.M. Hellerstein, Y. Ioannidis, H.V. Jagadish, T. Johnson, R. Ng, V. Poosala, K.A. Ross, and K. C. Sevcik. The New Jersey data reduction report. Data Engineering, 20(4):3--45, 1997.

.... 20 30 dimensions) a simple sequential scan usually performs better at higher dimensionalities [6, 43] To scale to higher dimensionalities, a commonly used approach is dimensionality reduction [20] This technique has been proposed for both multimedia retrieval [17, 36, 27, 42] and data mining ([18, 4, 21]) applications. The idea is to first reduce the dimensionality of the data and then index the reduced space using a multidimensional index structure [17] Most of the information in the dataset is condensed to a few dimensions (the first few principal components (PCs) by using principal component ....

D. Barbara, W. DuMouchel, C. Faloutsos, P. Haas, J. Hellerstein, Y. Ionnidis, H. Jagadish, T. Johnson, R. Ng, V. Poosala, K. Ross, and K. Sevcik. The new jersey data reduction report. Data Engineering, 20(4), 1997.

.... 20 30 dimensions) a simple sequential scan usually performs better at higher dimensionalities [6, 43] To scale to higher dimensionalities, a commonly used approach is dimensionality reduction [20] This technique has been proposed for both multimedia retrieval [17, 36, 27, 42] and data mining ([18, 4, 21]) applications. The idea is to first reduce the dimensionality of the data and then index the reduced space using a multidimensional index structure [17] Most of the information in the dataset is condensed to a few dimensions (the first few principal components (PCs) by using principal component ....

D. Barbara, W. DuMouchel, C. Faloutsos, P. Haas, J. Hellerstein, Y. Ionnidis, H. Jagadish, T. Johnson, R. Ng, V. Poosala, K. Ross, and K. Sevcik. The new jersey data reduction report. Data Engineering, 20(4), 1997.

....that would be found from fitting almost any smooth model to the larger data set. Data squashing can be seen as a form of lossy database compression. A significant body of recent work in the database literature has examined methods for the lossy compression of databases, and especially data cubes (Barbara 1997). However, data squashing has a somewhat different goal. Lossy data cube compression can be judged acceptable if aggregate queries over ranges of the data cube have a small error. Data squashing is acceptable if a different type of query has a small error, e.g. the fitting of statistical models. ....

Barbara, D. (1997). The New Jersey data reduction report. Bulletin on the Technical Committee on Data Engineering 20 (4), 3--45.

....group data of one attribute into buckets on the base of the data distribution and their frequency. Histograms approximate the frequency (typically by its average) and the values in each bucket. Clusters collect data that are similar to one another (see a description of clustering techniques in [1]) Distance models to define suitable clusters are required for different data types (medium) Access statistics are also usable for grouping data objects. Selections reduce the number of objects. Application characteristics and user needs can be expressed in reducing selections. In mobile ....

....3: Kinds of Data Reduction 3 Related Work Mobile approaches (like [2, 3, 7] are agree, that adaptation is a basic mobile concept. However, the consideration of reduction techniques as a special kind of adaptation in a comprehensible way and not only for images and video data, we are missing. [1], the Data Reduction Report, is a collection of different data reduction techniques. Data reduction, the authors believe, is a widely used technique in future database systems to support quick, but approximate answers from very large data sets in the context of data warehouse and data analysis ....

D. Barbara, W. DuMochel, C. Faloutsos, P.J. Haas, J.M. Hellerstein, Y. Ioannidis, H.V. Jagadish, T. Johnson, R. Ng, V. Poosala, K.A. Ross, K.C. Sevcik, "The New Jersey Data Reduction Report", Data Engineering, Vol.20 No.4, 1997;

....inserted or loaded into the tree. Bottom up methods, also called clustering methods, aim at identifying clusters embedded in data in order to reduce the search to clusters that potentially contain the nearest neighbor of the query. Several surveys provide background and analysis of these methods [1, 3, 15, 29]. Although these access methods generally work well for low dimensional spaces, their performance is known to degrade as the number of dimensions increases a phenomenon which has been termed the dimensional curse. This phenomenon has been reported for the R tree [7] the X tree [4] and the ....

....but also spherical MBRs, and a general class of clustering and partitioning methods. There exists a considerable number of reduction methods such as SVD, eigenvalue decomposition, wavelets, or Karhunen Lo eve transformation which can be used to decrease the effective dimensionality of a data set [1]. Faloutsos and Kamel [17] have shown that fractal dimensionality is a useful measure of the inherent dimensionality of a data set. We will further discuss this below. The indexability results of Hellerstein et al. 22] are based on data sets that can be seen as regular meshes of extension n in ....

[Article contains additional citation context not shown here]

D. Barbara, W. DuMouchel, C. Faloutsos, P. J. Haas, J. Hellerstein, Y. Ioannidis, H. Jagadish, T. Johnson, R. Ng, V. Poosala, K. Ross, and K. C. Sevcik. The New Jersey data reduction report. Data Engineering, 20(4):3-- 45, 1997.

....Comparison 4.1.1. MaxDiff and Modified MaxDiff Histograms Histograms approximate the data in one or more attributes of a relation by grouping attribute values into buckets and approximating the true attribute values and their frequencies based on summary statistics maintained in each bucket [3]. By replacing the frequencies with the measure attribute values, we can use histograms to approximate a data cube. Since the data cube is a multidimensional array, we concentrate on multidimensional histograms in our discussion. Muralikrishna and DeWitt [17] use an interesting spatial index ....

D. Barbara et al. The New Jersey data reduction report. Bulletin of the Technical Committee on Data Engineering, 20(4), 1997.

No context found.

D. Barbara, W. DuMouchel, C. Faloutsos, P. J. Haas, J. M. Hellerstein, Y. E. Ioannidis, H. Jagadish, T. Johnson, R. T. Ng, V. Poosala, K. A. Ross, and K. C. Sevcik. The New Jersey data reduction report. Data Engineering Bulletin, 20(4):3--45, 1997.

....state records are proportional in size to some (perhaps statistical) property of the data values in the partition. Many approximate aggregates proposed recently in the database literature are content sensitive. Examples of such aggregates include fixed width histograms, wavelets, and so on; see [3] for an overview of such functions. In summary, we have classified aggregates according to their state requirements, tolerance of loss, duplicate sensitivity, and monotonicity. We will refer back to this classification throughout the text, as these properties will determine the applicability of ....

D. Barbara, W. DuMouchel, C. Faloutsos, P. J. Haas, J. M. Hellerstein, Y. E. Ioannidis, H. Jagadish, T. Johnson, R. T. Ng, V. Poosala, K. A. Ross, and K. C. Sevcik. The New Jersey data reduction report. Data Engineering Bulletin, 20(4):3--45, 1997.

No context found.

D. Barbara, W. DuMouchel, C. Faloutsos, P. J. Haas, J. M. Hellerstein, Y. Ioannidis, H. V. Jagadish, T. Johnson, R. Ng, V. Poosala, K. A. Ross, and K. C. Sevcik. The New Jersey data reduction report. Bulletin of the Technical Committee on Data Engineering, 20(4):3--45, 1997.

....only index non empty cells (i.e. the data items that are in the database) can deal with sparseness. On the other hand just providing fast access to all selected data items does not suffice. Retrieving and aggregating each selected item onthe fly is still too slow for large data collections. In [2] it is suggested to augment indexes by aggregate data to exploit summary information for a data set. This idea up to now was used to support query cost estimation and approximate answers to queries [29] but to the best of our knowledge never for progressive feedback with absolute error bounds on ....

....the number of non empty cells in an arbitrary range of the data cube is a hard task. How detailed should the information about the data distribution be and how should it be organized for fast accesses and updates We will show how pCube answers these questions. The New Jersey Data Reduction Report [2] provides an overview of various techniques that can be useful for online aggregation. Barbara and Wu [4] describe a technique to use loglinear models for compressing the data cube and obtaining approximate answers. For the technique to be efficient, dense clusters in the data cube have to be ....

D. Barbara et al. The new jersey data reduction report. Data Engineering Bulletin, 20(4), 1997.

No context found.

D. Barbara, W. DuMouchel, C. Faloutsos, P. J. Haas, J. M. Hellerstein, Y. E. Ioannidis, H. V. Jagadish, T. Johnson, R. T. Ng, V. Poosala, K. A. Ross, and K. C. Sevcik. The New Jersey Data Reduction Report. IEEE Data Engineering Bulletin, 20(4), December 1997.

No context found.

D. Barbara, W. DuMouchel, C. Faloutsos, P. J. Haas, J. M. Hellerstein, Y. Ioannidis, H. V. Jagadish, T. Johnson, R. Ng, V. Poosala, K. A. Ross, and K. C. Sevcik. The New Jersey data reduction report. Bulletin of the Technical Committee on Data Engineering, 20(4):3--45, 1997.

No context found.

Daniel Barbara, WilliamDuMouchel, Christos Faloutsos, Peter J. Haas, Joseph M. Hellerstein, Yannis E. Ioannidis, H. V. Jagadish, Theodore Johnson, Raymond T. Ng, Viswanath Poosala, Kenneth A. Ross, and Kenneth C. Sevcik. The New Jersey Data Reduction Report. IEEE Data Engineering Bulletin, 20(4), December 1997. 21

No context found.

D. Barbara, W. DuMouchel, C. Faloutsos, P. J. Haas, J. M. Hellerstein, Y. Ioannidis, H. V. Jagadish, T. Johnson, R. Ng, V. Poosala, K. A. Ross, and K. C. Sevcik. The New Jersey data reduction report. Bulletin of the Technical Committee on Data Engineering, 20(4):3--45, 1997.

....reasonable accuracy. 3 Overview of Cluster based Database Selection Clustering refers to the grouping of database records based on the degrees of similarity between the records. Clustering has been used in many fields, such as information retrieval (IR) 19, 4, 14] data mining, data reduction[1], etc. In order to route queries to a set of databases each with multiple text attributes, the content of each databases has to be summarized properly. Nevertheless, as the databases contain wide range of information, direct summarization of their content may result in inaccurate summary ....

D. Barbara, W. DuMouchel, C. Faloutsos, P.J. Haas, J.M. Hellerstein, Y. Ioannidis, H.V. Jagadish, T. Johnson, R. Ng., V. Poosala, K.A. Ross, and K.C. Sevcik. The new jersey data reduction report. Bulletin of the Technical Committee on Data Engineering, 20(4), December 1997.

No context found.

D. Barbara, W. DuMouchel, C. Faloutsos, P. Hass, J. M. Hellerstein, Y. Ioannidis, H. Jagadish, T. Johnson, R. Ng, V. Poosala, K. Ross, and K. Sevcik. The new jersey data reduction report. Bulletin of the Technical Committee on Data Engineering, 20(4):3--45, December 1997.

No context found.

Barbara, D., DuMouchel, W., Faloutsos, C., Haas, P. J., Hellerstein, J. M., Ioannidis, Y., Jagadish, H. V., Johnson, T., Ng, R., Poosala, V., Ross, K. A., Sevcik, K. C., The New Jersey data reduction report, Bulletin of the Technical Committee on Data Engineering 20, 4, 3-45, 1997.

No context found.

Barbara et al. "The new jersey data reduction report". Data Engineering Bulletin, September 1996

No context found.

Barbara, D., DuMouchel, W., Faloutsos, C., Haas, P.J., Hellerstein, J.M., Ioannidis, Y.E., Jagadish, H.V., Johnson, T., Ng, R.T., Poosala, V., Ross, K.A., Sevcik, K.C.: The new jersey data reduction report. IEEE Data Engineering Bulletin 20 (1997) 3--45

No context found.

Daniel Barbara, William DuMouchel, Christos Faloutsos, Peter J. Haas, Joseph M. Hellerstein, Yannis E. Ionnidis, H.V. Jagadish, Theodore Johnson, Raymond T. Ng, and Viswanath Poosala. The new jersey data reduction report. Data Engineering Bulletin, 20(4), 1997.

No context found.

Barbara, D., DuMouchel, W., Faloutsos, C., Haas, P.J., Hellerstein, J.M., Ioannidis, Y.E., Jagadish, H.V., Johnson, T., Ng, R.T., Poosala, V., Ross, K.A., Sevcik, K.C.: The new jersey data reduction report. IEEE Data Engineering Bulletin 20 (1997) 3--45

No context found.

Barbara et al. "The new jersey data reduction report". Data Engineering Bulletin, September 1996

No context found.

D. Barbara, W. DuMouchel, C. Faloutsos, P. J. Haas, J. M. Hellerstein, Y. E. Ioannidis, H. V. Jagadish, T. Johnson, R. T. Ng, V. Poosala, K. A. Ross, and K. C. Sevcik. The New Jersey Data Reduction Report. IEEE Data Engineering Bulletin, 20(4):3-45, 1997.

No context found.

Barbara D, DuMouchel W, Faloutsos C, Haas PJ, Hellerstein JM, Ioannidis YE, Jagadish HV, Johnson T, Ng RT, Poosala V, Ross KA, Sevcik KC (1997) The New Jersey Data Reduction Report. IEEE Data Eng Bull 20(4):3--45

No context found.

D. Barbara, W. DuMouchel, C. Faloutsos, P. Haas, J. Hellerstein, Y. Ioannidis, H. Jagadish, T. Johnson, R. Ng, V. Poosala, K. Ross, and K. Sevcik. The New Jersey data reduction report. Data Engineering Bulletin, 20(4), 1997.

No context found.

Daniel Barbara, William DuMouchel, Christos Faloutsos, Peter J. Haas, Joseph M. Hellerstein, Yannis Ioannidis, H.V. Jagadish, Theodore Johnson, Raymond Ng, Viswanath Poosala, Kenneth A. Ross, and Kenneth C. Sevcik. The New Jersey Data Reduction Report. Data Engineering Bulletin, 20(4):3--45, December 1997.

No context found.

Barbara, D., DuMouchel, W., Faloutsos, C., Haas, P. J., Hellerstein, J. M., Ioannidis, Y. E., Jagadish, H. V., Johnson, T., Ng, R. T., Poosala, 169 V., Ross, K. A., and Sevcik, K. C., "The new jersey data reduction report," IEEE Data Engineering Bulletin, vol. 20, no. 4, pp. 3--45, 1997.

No context found.

D. Barbara, C. Faloutsos, J. Hellerstein, Y. Ioannidis, H.V. Jagadish, T. Johnson, R. Ng, V. Poosala, K. Ross, and K.C. Sevcik, "The New Jersey Data Reduction Report," Data Eng. Bull., Sept. 1996.

No context found.

D. Barbara, W. DuMouchel, C. Faloutsos, P. J. Haas, J. M. Hellerstein, Y. E. Ioannidis, H. V. Jagadish, T. Johnson, R. T. Ng, V. Poosala, K. A. Ross, and K. C. Sevcik. The New Jersey Data Reduction Report. IEEE Data Engineering Bulletin, 20(4):3-45, 1997.

No context found.

Barbara D., et al.: The New Jersey Data Reduction Report. Data Engineering Bulletin 20:4 (1997) 3-45

No context found.

D. Barbar'a, W. DuMouchel, C. Faloutsos, P. J. Haas, J. M. Hellerstein, Y. Ioannidis, H. V. Jagadish, T. Johnson, R. Ng, V. Poosala, K. A. Ross, and K. C. Sevcik. The New Jersey data reduction report. Bulletin of the Technical Committee on Data Engineering, 20(4):3--45, 1997.

No context found.

D. Barbara, C. Faloutsos, J. Hellerstein, Y. Ioannidis, H. V. Jagadish, T. Johnson, R. Ng, V. Poosala, K. Ross, and K.C. Sevcik. The new jersey data reduction report. Data Engineering Bulletin, September 1996.

No context found.

D. Barbara, W. DuMouchel, C. Faloutsos, P. J. Haas, J. M. Hellerstein, Y. Ioannidis, H. V. Jagadish, T. Johnson, R. Ng, , V. Poosala, K. A. Ross, and K. C. Sevcik. The New Jersey data reduction report. Bulletin of the Technical Committee on Data Engineering, 20(4), 1997.

No context found.

Daniel Barbara, William DuMouchel, Christos Faloutsos, Peter J. Haas, Joseph M. Hellerstein, Yannis E. Ionnidis, H.V. Jagadish, Theodore Johnson, Raymond T. Ng, and Viswanath Poosala. The new jersey data reduction report. Data Engineering Bulletin, 20(4), 1997.

No context found.

D. Barbara, W. DuMouchel, C. Faloutsos, P. J. Haas, J. M. Hellerstein, Y. Ioannidis, H. V. Jagadish, T. Johnson, R.Ng, V. Poosala, and K. Sevcik. The New Jersey Data Reduction Report. Data Engrg. Bull., 20:3--45, Dec. 1997.

No context found.

D. Barbara, W. DuMouchel, C. Faloutsos, P. J. Haas, J. M. Hellerstein, Y. Ioannidis, H. V. Jagadish, T. Johnson, R. Ng, V. Poosala, K. A. Ross, and K. C. Sevcik. The New Jersey data reduction report. Bulletin of the Technical Committee on Data Engineering, 20(4):3-- 45, 1997.

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