Data processing: database and file management or data structures – Database design – Data structure types
Reexamination Certificate
2001-06-18
2004-07-13
Alam, Shahid (Department: 2172)
Data processing: database and file management or data structures
Database design
Data structure types
C707S793000, C707S793000
Reexamination Certificate
active
06763349
ABSTRACT:
BACKGROUND OF THE INVENTION
The present invention refers to a dynamic taxonomy process for browsing and retrieving information in large heterogeneous data bases.
Information retrieval on this type of data bases (for example those available on the Internet) is nowadays a slow task, sometimes impossible to realize due to the enormous amount of data to be analyzed, and that can be implemented with difficulty with the currently available tools. The following documents deal with the prior art in this field: Hearst M. et al: “Cat-a-cone: an interactive interface for specifying searched and viewing retrieval results using a large category hierarchy,” Annual International ACM-SIGIR Conference on Research and Development in Information Retrieval, US, New York, N.Y.: ACM, 1997, pages 246-255; EP-A-0 694 829 (XEROX Corp.); U.S. Pat. No. 5,644,740 (Kiuchi Itsuko); Gert Schmeltz Pedersen: “A browser for bibliographic information retrieval, based on an application of lattice theory,” Proceedings of the Annual International ACM SIGIR Conference on Research and Development in Information Retrieval, US, New York, ACM, vol. CONF., 16, 1993, pages 270-279; and Story G. et al: “The Rightpages image-based electronic library for alerting and browsing,” Computer, US, IEEE Computer Society, Long Beach, Calif., US, vol. 25, no. 9, 1 Sep. 1992, pages 17-25.
Dynamic taxonomies are a model to conceptually describe and access large heterogeneous information bases composed of texts, data, images and other multimedia documents.
The following documents deal with prior art in this field: Hearst M. et al: ‘Cat-a-cone: an interactive interface for specifying searched and viewing retrieval results using a large category hierarchy’, Annual International ACM-SIGIR Conference on Research and Development in Information Retrieval, US, New York, N.Y.: ACM, 1997, pages 246-255; EP-A-0 694 829 (XEROX Corp.); U.S. Pat. No. 5,644,740 (Kiuchi Itsuko); Gert Schmeltz Pedersen: ‘A browser for bibliographic information retrieval, based on an application of lattice theory’, Proceedings of the Annual International ACM SIGIR Conference on Research and Development in Information Retrieval, US, New York, ACM, vol. CONF., 16, 1993, pages 270-279; and Story G. et al: ‘The Rightpages image-based electronic library for alerting and browsing”, Computer, US, IEEE Computer Society, Long Beach, Calif., US, vol. 25, no. 9, 1 Sep. 1992, pages 17-25.
As disclosed in Hearst, a dynamic taxonomy is basically a IS-A hierarchy of concepts, going from the most general (topmost) to the most specific. A concept may have several fathers. This is a conceptual schema of the information base, i.e. the “intension”. Documents can be freely classified under different concepts at different level of abstraction (this is the “extension”). A specific document is generally classified under several concepts.
Dynamic taxonomies enforce the IS-A relationship by containment, i.e. the documents classified under a concept C are the deep extension of C, i.e. the recursive union of all the documents classified under C and under each descendant C′ of C.
In a dynamic taxonomy, concepts can be composed through classical boolean operations. In addition, any set S of documents in the universe of discourse U (defined as the set of all documents classified in the taxonomy) can be represented by a reduced taxonomy. S may be synthesized either by boolean expressions on concepts or by any other retrieval method (e.g. “information retrieval”). The reduced taxonomy is derived from the original taxonomy by pruning the concepts (nodes) under which no document d in S is classified.
A new visual query/browsing approach is supported by dynamic taxonomies. The user is initially presented with the complete taxonomy. He/she can then refine the result by selecting a subset of interest. Refinement is done by selecting concepts in the taxonomy and combining them through boolean operations. She/he will then be presented with a reduced taxonomy for the selected set of documents, which can be iteratively further refined.
The invention described here covers the following aspects of dynamic taxonomies:
1. additional operations;
2. abstract storage structures and operations on such structures for the intension and the extension;
3. physical storage structures, architecture and implementation of operations;
4. definition, use and implementation of virtual concepts;
5. definition, use and implementation of time-varying concepts;
6. binding a dynamic taxonomy to a database system;
7. using dynamic taxonomies to represent user profiles of interest and implementation of user alert for new interesting documents based on such profiles of interest.
The above and other objects and advantages of the invention, as will appear from the following description, are obtained by a dynamic taxonomy process as claimed in claim
1
. Preferred embodiments and non-trivial variations of the present invention are claimed in the dependent Claim.
REFERENCES:
patent: 5644740 (1997-07-01), Kiuchi
patent: 0 694 829 (1996-01-01), None
M. A. Hearst, et al., Annual International ACM-SIGIR Conference on Research and Development in Information Retrieval, pp. 246-255, “CAT-A-CONE: An Interactive Interface for Specifying Searched and Viewing Retrieval Results Using a Large Category Hierarchy”, 1997.
G. Schmeltz Pedersen, Proceedings of the Annual International ACM SIGIR Conference on Research and Development in Information Retrieval, vol. 16, pp. 270-279, “A Browser for Bibliographic Information Retrieval, Based on an Application of Lattice Theory”, 1993.
G. A. Story, et al., Computer, US, IEEE Computer Society, vol. 25, No. 9, pp. 17-25, “The Rightpages Image-Based Electronic Library for Alerting and Browsing”, Sep. 1, 1992.
Alam Shahid
Oblon & Spivak, McClelland, Maier & Neustadt P.C.
Woo Isaac
LandOfFree
Dynamic taxonomy process for browsing and retrieving... 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 taxonomy process for browsing and retrieving..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Dynamic taxonomy process for browsing and retrieving... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3248986