Data processing: database and file management or data structures – Database design – Data structure types
Reexamination Certificate
2002-11-14
2004-12-28
Corrielus, Jean M. (Department: 2172)
Data processing: database and file management or data structures
Database design
Data structure types
C707S793000, C707S793000
Reexamination Certificate
active
06836777
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The invention disclosed herein relates generally to database applications, and more particularly to a system and method for constructing generic analytical database applications through the automated creation of metadata to establish an application structure controlling the availability and operability of individual applications.
2. Description of the Background
Market analysis is a key tool in customer acquisition and retention. Performing a detailed market analysis of both potential product or service purchasers and existing purchasers, and from such analysis obtaining an understanding of customer behavior, product success, marketing success, and other business factors enables a supplier to make critical decisions regarding how they conduct business in order to maximize both new customer acquisition and the retention of existing customers. However, performing such a detailed analysis in order to obtain an accurate depiction of customer behavior can be a daunting task.
Many product and service providers maintain records of particular customer data, including demographic information, purchasing characteristics, etc., in addition to maintaining detailed records of products, pricing, and other business information. The collections of such data in electronic form can become enormous for large entities and attempting to digest such information into a logical structure from which one might be able to deduce customer trends and characteristics can be a highly technologically challenging task.
In order to analyze such large volumes of information to in turn make reasoned deductions about customer characteristics, product sales, and other business-related information, automated data mining tools have been implemented which attempt to identify patterns in data which might not be self-evident in a review of the data by a human analyst. Such data mining tools traditionally seek to extract patterns that exist in large volumes of data in order to forecast future expectations, whether of customer actions, future profit, inventory needs, scheduling requirements, or any other trend or characteristic that a planner may wish to forecast.
Unfortunately, however, the implementation of a successful data mining strategy often requires persons having highly technical expertise to develop accurate data mining profiles that might model such future expectations. Before a user may know what questions may be asked, the user must know what data is available to answer his questions. Likewise, in order to ask a particular question for which the data might be able to provide an answer, the user must know how to structure the question, i.e., have knowledge of the query language required to ask the question. The fact that database fields, customers, and analytical database applications all have distinct labels for the same conceptual piece of data further exemplifies the need for expert-level knowledge of the data structure. Given such labeling discrepancy, while a particular non-expert user may know in plain English the question he wishes to have answered, he lacks the expert knowledge of how that data is arranged or how the applications might extract and manipulate that data, and thus whether and how the data might provide that answer. While the creation of metadata describing how the data is arranged may to a limited extent alleviate some of the problems arising from such labeling discrepancy, the creation of metadata itself can also be a daunting task.
The development of a successful data mining strategy traditionally involves six distinct phases, as follows: (i) obtaining an understanding of the project objectives and requirements from a business perspective, then converting this knowledge into a data mining problem definition and a preliminary plan designed to achieve the objectives; (ii) initial data collection and activities to become familiar with the data, identify data quality problems, discover first insights into the data, and detect interesting subsets to form hypotheses for hidden information; (iii) constructing the final dataset to be used by the predictive modeling tools from the initial raw data, including categorizing continuous variables, eliminating outlying data values, joining related data sources, and extracting a sample of the result; (iv) selecting and applying various modeling techniques, and calibrating their parameters to optimal values, which often involves returning to the data preparation phase; (v) evaluating the model to be certain it properly achieves the business objectives and meets accuracy requirements; and (vi) deployment of the model, in which analysts consider the application of the model to new inputs, consider the application that will eventually receive the scored prospects, and consider the operational issues surrounding the ongoing execution of the model. Such a process often involves up to six months of expert time to develop a single model, with the business understanding, data understanding, and data preparation phases consuming the largest portion of this effort.
Obviously, it would be highly advantageous to provide a means by which the business understanding, data understanding, and data preparation phases could be implemented in less time so as to provide timely analysis, and without the need for such expert technical intervention.
Moreover, in order to provide adaptive applications that provide the greatest possible utility to users, it is necessary that the applications themselves be able to adapt to the data that is available, such that all application functions that are supported by the available data are readily available to the user, while application functions that are not supported are not available to the user.
SUMMARY OF THE INVENTION
It is, therefore, an object of the present invention to provide a method and system for constructing generic analytical database applications which avoids the disadvantages of the prior art.
In accordance with the above objects, a method and system for constructing generic analytical database applications is provided which allows users to access application functions which are enabled depending upon the data stored in databases, as reflected by metadata.
The metadata connects user terminology, application terminology, and database terminology in a common framework. The system of the instant invention allows the construction of analytical database applications having broad adaptability in terms of the database layout and business functions supported. The applications operate against an abstract business model, with all components (i.e., “roles”) optional. Application functions presented to the user depend on the configuration of roles which are present in the metadata. Databases are described in the metadata at an entity relationship level, and are annotated with semantic business roles, as well as with a wide variety of user preferences for naming and presentation. The end-user sees neither database terminology (SQL names) nor application terminology (roles). Rather, they see things in relevant business terms, and they only see applications and functions which are viable given the metadata. If there is a database in place, much of the metadata can be guessed automatically (e.g., table associations or roles). This means introduction of this technology provides a fast start, followed by incremental refinement which surfaces new application functions and more familiar user terminology.
The application environment of the instant invention seeks to provide a suite of applications which may be implemented across as many database scenarios as possible, and thus creates generic applications which use rich metadata to adapt to an existing data environment. The application environment of the instant invention provides a variety of information and services to the individual database applications, such as information on the user (e.g., user preferences, user privileges, etc.), and information on the data visible to that user (i.e., metadata describing the information both in
Corrielus Jean M.
NCR Corporation
Stover James M.
Whiteford, Talyor & Preston LLP
LandOfFree
System and method for constructing generic analytical... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with System and method for constructing generic analytical..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and System and method for constructing generic analytical... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3327179