Dynamic recursive build for multidimensional databases and...

Details Dynamic recursive build for multidimensional databases and... Dynamic recursive build for multidimensional databases and...

1999-12-13

2002-06-11

Coby, Frantz (Department: 2171)

Data processing: database and file management or data structures

Database design

Data structure types

C707S793000, C707S793000, C707S793000, C707S793000

Reexamination Certificate

active

06405208

ABSTRACT:

BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates generally to electronic databases. More particularly, the present invention relates to storing and retrieving data from multidimensional databases.
2. Description of the Related Art
One general category of application software is often referred to as a database management program or simply a database application. Encompassed within this general category are database systems referred to commercially as multidimensional databases or, in technical discussions, as Online Analytical Processing (OLAP) data stores. The OLAP paradigm is described in the white paper entitled “Providing OLAP (On-line Analytical Processing) to User-Analysts: An IT Mandate” by E. F. Codd, S. B. Codd, and C. T. Salley published by Codd & Date, Inc., and incorporated by reference herein for all purposes.
Typically, a multidimensional database stores and organizes data in a way that better reflects how a user would want to view the data than is possible in a spreadsheet or relational database file. Multidimensional databases are better suited generally to handle applications with large volumes of numeric data and that require calculations on numeric data, such as business analysis and forecasting.
A dimension within multidimensional data is typically a basic categorical definition of data in a database outline (discussed in greater detail below). A multidimensional database can contain several dimensions thereby allowing analysis of a large volume of data from multiple viewpoints or perspectives. Thus, a dimension can also be described as a perspective or view of a specific dataset. A different view of the same data is referred to as an alternative dimension. A data management system that supports simultaneous, alternative views of datasets is said to be multidimensional. Using a business application as an example, dimensions are items such as TIME, ACCOUNTS, PRODUCT LINES, MARKETS, DIVISIONS, and so on. Within each dimension, there is typically a consolidation or other relationship between items.
Information in a database can be stored and maintained in various data structures. To facilitate discussion,
FIG. 1A
is a simplified representation of an exemplary database table
100
. Data table
100
can be used to store information relating to different products that are available for sale in a department store. As shown in
FIG. 1A
, data table
100
can include a product ID, a product category, a product sub-category, and a product level.
Each product category, e.g., “clothing”, “men's clothing”, etc. can be assigned a unique product ID. Information relating to a particular product category can be accessed using its product ID. For example, product ID “
2
” in table
100
uniquely identifies “informal wear”. Thus, by looking up product ID “
2
” in table
100
, other information such as the fact that product categories “jeans” and “t-shirts” are sub-categories of “informal wear” can be ascertained.
Other information relating to product categories can be maintained in another table, sales table
110
of FIG.
1
B. As shown in
FIG. 1B
, the sales table
110
includes specific information relating to sale of products, e.g., “retail ID”, “current price”, and “number sold this year.” For example, Information related to the “current price” for “Jeans” can be obtained by accessing the table
110
using “
3
” as the product ID and looking up the appropriate column, i.e., “current price”.
FIG. 1C
depicts another informational aspect (dimension) related to a department store. Namely, the organizational arrangement (“topology”) of the department store. Similar to product table
100
, each particular region shown in
FIG. 1C
, e.g., “U.S.”, “West Coast” can be assigned a unique regional ID. Information relating to topology of a particular region can be accessed using the regional ID that uniquely identifies that region.
As is known in the art, the Information maintained in a database, e.g., information in data tables shown in
FIGS. 1A-1C
, can be used to solve various analytical problems. By way of example, the department store can use the information kept in its database to solve problems of having to keep track of inventory, sales, employee records and salaries, and so on. In order to address an analytical problem, it is often necessary to combine information maintained in several database tables. In addition, an analytical problem may be related to one or more dimensions of data. By way of example, the user may wish to now the 3-rd quarter sales for a product category and all its sub-categories (e.g., “informal wear” with subcategories of “jeans” and “T-shirts”) for a regional store and its sub-regions (e.g., San Diego with its sub-region of La Jolla). This problem involves at least two dimensions related to the general category of“sales”, namely, the dimensions of “products” and “topology”. Thus, in that example, the information in tables shown in
FIGS. 1A-C
need to be combined (merged) to adequately generate the desired sale reports.
Moreover, it is useful to have the solution organized and presented in a way that better reflects how a human would want to view the data. To elaborate, there is an implicit hierarchical (parent-child) relationship between the product categories of database table
100
. For example, “informal wear” is a child of “clothing” and a parent of both “jeans” and “T-shirts”. The hierarchical relationship present in table
100
is illustrated in
FIG. 2A
, where, for example, “Informal wear” is represented as a child of “men's clothing” and as a parent of both “jeans” and “T-shirts”. The graphical presentation illustrated in
FIG. 2A
is better suited for human perception and comprehension. This is evident from a quick comparison of
FIG. 1A
to FIG.
2
A. As another example,
FIG. 2B
illustrates the hierarchical relationship present in FIG.
1
C. Again, the graphical representation of
FIG. 2B
is a more desirable presentation to a user.
As mentioned earlier, the multidimensional databases have the ability to present a user with several different views (dimensions) of data. To facilitate understanding, a multidimensional solution provided by a multidimensional database can be represented by a multidimensional structure, e.g., a cube
120
of
FIG. 2C
wherein each side of the cube
120
can present the user with a different dimension of data. For example, sides
122
and
124
of the cube
120
can contain the hierarchical relationships depicted in
FIGS. 2A and 2B
respectively.
Recently, there have been significant developments in the area of multidimensional databases. However, primarily “brute force” approaches have been used to generate multidimensional outputs without much regard to overall cost and efficiency. As is known in the field, brute force approaches generally require making several passes through relevant tables in a database, of which there maybe many, to ultimately generate an appropriate multidimensional output.
By way of example, in order to generate a multidimensional output that adequately addresses the problem of determining the 3-rd quarter sales for a product category and its sub-categories for a regional store and its sub-regions, several passes through tables shown in
FIGS. 1A-1C
would have to be made. Consequently, brute force approaches are inefficient and expensive with respect to system resources. In addition, since making several passes through a database table requires significant amount of time, brute force approaches are not suitable for multidimensional databases where it is desirable to quickly present the user with several dimensions of data on demand (“on the fly”).
Another approach for generating multidimensional outputs is to “normalize” data. Normalization of data generally puts data in a format that is more readily suitable for generation of multidimensional output. Database table
130
of
FIG. 2D
illustrates how the hierarchical relationship present in table
1
A (also illustrated in the tree representation of
FIG. 2A
) may be normalized. As illu

Affiliated with

Also associated with

Rating

Dynamic recursive build for multidimensional databases and... does not yet have a rating. At this time, there are no reviews or comments for this patent.

If you have personal experience with Dynamic recursive build for multidimensional databases and..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Dynamic recursive build for multidimensional databases and... will most certainly appreciate the feedback.

Rate now

Comments { 0 }

Profile ID: LFUS-PAI-O-2947673