Data processing: database and file management or data structures – Database design – Data structure types
Reexamination Certificate
2001-07-05
2003-10-21
Mizrahi, Diane D. (Department: 2175)
Data processing: database and file management or data structures
Database design
Data structure types
Reexamination Certificate
active
06636862
ABSTRACT:
FIELD OF THE INVENTION
This application relates to a method and system for the dynamic analysis of data, especially data represented in distinct matrices, for example, X, Y and Z data matrices. If two data matrices X and Y are present in which corresponding rows of X and Y each refer to the same underlying object, a relationship can developed between the X and Y data matrices, which allows the method and system of the present invention to predict responses in Y on the basis of inputted X-data. And, if a third data matrix Z is present in which corresponding columns of Y and rows of Z each refer to the same underlying object, a relationship can developed between the X, Y and Z data matrices, which allows the method and system of the present invention to link X with Z through Y.
BACKGROUND OF THE INVENTION
Again, this application relates to a method and system for the dynamic analysis of data, for example, data related to consumer choice modeling or quality control programs.
Advances in computerization and other technologies have greatly facilitated the collection of data. However, once data has been collected, entities are often faced with new problems and challenges related to the analysis and use of the accumulated data. If there is not an effective method and system for storing, updating, and analyzing the accumulated data, its value is minimal. Such problems and challenges are compounded by the desire of entities to gain a better understanding of the complex relationships inherent in the data, and the further desire of entities to access and analyze such relationships, particularly in the real-time world of global computing and the Internet.
In response, various “data mining” or “knowledge extraction” methods and systems have been developed to assist in extracting conclusions and answers from accumulated data. Specifically, such data mining and knowledge extraction methods and systems use approaches for cleaning data sets, coding data values, deriving secondary attributes, developing relationships and classes, and reporting. In this regard, a wide range of statistical and artificial intelligence methods is used in connection with such approaches. Despite the approach or combination of approaches used, the goal is ultimately to provide useful conclusions through an analysis of the accumulated data.
Perhaps one of the best examples of the use of data mining or knowledge extraction is in the consumer products industry. As can be expected, competition is fierce in the manufacture and marketing of consumer products. Manufacturers of consumer products expend great resources in identifying the preferences and attitudes of their targeted markets. In so identifying these preferences and attitudes, manufacturers of consumer products often accumulate vast amounts of raw data. As discussed above, without an appropriate analysis of the data, its value is minimal.
Current data mining and knowledge extraction techniques, however, are not always effective. One of the primary problems is the pace at which consumer preferences and attitudes shift and change. Many of the prior art analysis tools can not rapidly adapt to changing market conditions. For example, a manufacturer may accumulate twelve months worth of raw data regarding consumer preferences, create a consumer preference model through an analysis of that data, and then package and market its new product accordingly. However, since the manufacturer did not account for the dynamics of the marketplace, the shifting and changing of consumer preferences, its model may prove ultimately ineffective.
Furthermore, prior art approaches commonly are designed to can relate two types of at objects (i.e., row and column objects) on the basis of some measure, e.g., correspondence analysis, principal components analysis, and factor analysis. While such approaches assist in developing an understanding of data relationships or associations, they do not allow for the incorporation of the respective independent qualities of objects into models, qualities that that define predictive (e.g., cause and effect) relationships between independent objects. Similarly, prior art approaches do not allow for the grouping of segments of objects of one type with respect to some measure on another set of objects of another type, nor do they allow for the use of the qualities of the respective objects to build predictive models relating the qualities of the two respective types of objects.
It is therefore a paramount object of the present invention to provide a method and system for the dynamic analysis of data that overcomes the inefficiencies and shortcomings of prior art methods and systems, a method and system that allow for the incorporation of the respective independent qualities of objects into models, qualities that that define predictive relationships between independent objects.
This and other objects and advantages of the present invention will become apparent upon a reading of the following description.
SUMMARY OF THE INVENTION
The present invention is a method and system for the dynamic analysis of data that is comprised of a series of computational steps achieved through the use of a digital computer program. Specifically, there are three phases of the preferred dynamic data analysis. In the first phase, a cluster analysis or clustering sequence is performed on a data matrix Y in order to segment the data into appropriate clusters for subsequent computational analysis. The second phase is a modeling phase. If a second data matrix X is present such that corresponding rows of X and Y each refer to the same underlying object, in the second phase, a relationship can developed between the X and Y data matrices, which allows the method and system of the present invention to predict responses in Y on the basis of inputted X-data. The third phase is also a modeling phase. If a third data matrix Z is present in which corresponding columns of Y and row of Z each refer to the same underlying object, a relationship can developed between the X Y and Z data matrices, which allows the method and system of the present invention to link X with Z through Y.
REFERENCES:
patent: 5325298 (1994-06-01), Gallant
patent: 5442778 (1995-08-01), Pedersen et al.
patent: 5945675 (1999-08-01), Malins
patent: 6012058 (2000-01-01), Fayyad et al.
patent: 6256633 (2001-07-01), Dharap
patent: 6263334 (2001-07-01), Fayyad et al.
patent: 2215891 (1989-09-01), None
Data Clustering: A Review By Jain et al., ACM Computing Surveys, vol. 31, No. 3, Sep. 1999.*
Lee, Daeryong, et al., Modified K-Means Algorithm for Vector Quantizer Design, IEEE Signal PRocessing Letters, Jan. 1997, p. 2-4, vol. 4, Issue 1.
Newton, S.C., et al., Adaptive Fuzzy Leader Clustering of Complex Data Sets in Recognition, Neural Networks, IEEE Transactions On, Sep. 1992, vol. 3 Issue 5, p. 794-800.
Rose, K., et al., Hierarchical Clustering Using Determistic Annealing, Neural Networks, 1992, IJCNN, Internation Joint Conference on, Jun. 1992, vol. 4, p. 85-90.
Lee, Daeryong et al., Generalization of the Codebook Updating Method of the k-Means Algorithm, Communications 1999, APCC/DECC '99 Fifth Asia Conference on . . . And Forth Optoelectronics and Communications, Oct. 1999, vol. 1, p. 566-569.
Zamir, Owen et al., Web document clustering: a feasibility demonstration, Proceedings of the 21st annual international ACM SIGIR conference on Research and development in information retrieval, Aug. 24-28, 1998, p. 46-54, Melbourne, Australia.
Anick, Peter G. et al., Exploiting clustering and phrases for context-based information retrieval, Proceedings of the 20th annual international ACM SIGIR conference on Research and development in information retrieval, Jul. 27-31 , 1997, p. 314-323, Philadelphia, Pennsylvania, USA.
International Search Report for International Application No. PCT/US01/21663, mailed Nov. 15, 2001.
Dayal, Bhupinder, S. and Macgregor, John F.; Recursive exponentially weighted PLS and its application to adaptive control and prediction: J. Proc. Cont. vol. 7, No. 3, pp. 169-179 (1997) (Elsevier Science Ltd., Gre
Lundahl David S.
Plaehn David C.
Camo, Inc.
Mizrahi Diane D.
Mofiz Apu M
Nagle, Jr. David W.
Stites & Harbison PLLC
LandOfFree
Method and system for the dynamic analysis of data does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method and system for the dynamic analysis of data, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method and system for the dynamic analysis of data will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3145728