Relational database system containing a multidimensional hierachical model of interrelated subject categories with recognition capabilities

• US 5,721,910 A
• Filed: 06/04/1996
• Issued: 02/24/1998
• Est. Priority Date: 06/04/1996
• Status: Expired due to Term

• Current Assignee: Exxon Research and Engineering Company (Exxon Mobil Corporation)
• Original Assignee: Exxon Research and Engineering Company (Exxon Mobil Corporation)
• Inventors: Unger, Sandra S., Pagnucco, Sibylle, Cohen, Roger W., Fiato, Rocco A.
• Primary Examiner(s): AMSBURY, WAYNE P
• Application Number:
  US08/655,262
• Time in Patent Office: 630 Days
• Field of Search: 395/611, 395/614, 395/615
• US Class Current: 707/758
• CPC Class Codes: G06F 16/30 of unstructured textual data document management systems G06F16/93
• CPC Class Codes: Y10S 707/929 Docketing
• CPC Class Codes: Y10S 707/954 Relational
• CPC Class Codes: Y10S 707/956 Hierarchical
• CPC Class Codes: Y10S 707/957 Multidimensional
• CPC Class Codes: Y10S 707/99945 Object-oriented database structure processing

Abstract

The present invention is a database and a method of producing that database which can be used to determine the meaning of scientific or technical documents, such as patents and/or technical or scientific publications and/or abstracts of these patents or publications, and to assign the technical documents to one or more scientific or technical categories within a multidimensional hierarchical model which reflects the business, scientific or technical interests of a business, scientific or technical entity or specialty.

Claims

1. A computer readable medium comprising a database system, including stored data of scientific or technical documents, such as patents, technical or scientific publications, abstracts of these patents or publications, and associated bibliographic and technical classification data, or combinations thereof such that said scientific and technical documents are assigned to one or more scientific or technical categories (category assignments) within a multidimensional hierarchical model of a business, scientific or technical entity or specialty, and said category assignments are stored in a relational database wherein said category assignments are used to identify one or more patterns, trends and/or discontinuities based on a population analysis of the categories.

2. The database system of claim 1 where the stored data and category assignments can be displayed on a computerized graphical device in the form of multidimensional tabular or graphical displays which permit the observation of trends and discontinuities within the business, scientific or technical entity or specialty being modeled.

3. The database of claim 1 wherein said stored data includes patent abstracts, patent identifiers, technical index tenths, patent classifications, patent claims, graphic images, chemical structures, and all or portions of the text of the patent or published patent application.

4. The database of claim 1 wherein said stored data includes the technical abstracts, document identifiers, technical index terms, graphic images, chemical structures, and all or portions of the text of the scientific or technical publication.

5. The database of claim 1 wherein said stored data includes a fractional contribution matrix which stores the cumulative opinion of a team of technical or scientific experts on the fractional contribution of each category to a set of higher-level (more abstract) concepts within the hierarchical model.

6. The database of claim 5 wherein a set of higher-level (more abstract) concepts are related to each stored category by a mathematical relationship.

7. The database system of claim 1 where the database can also store the claims of the corresponding U.S. patents and the claims of the corresponding European patents.

11. A method as of claim 7 where multiple-entity fields are parsed to create normalized, single-entity fields in a separate table including a unique identifier for each patent or technical document.

12. A method as of claim 7 where the expert technical search strategies are stored in a database and are applied against technical index terms, text, and/or abstract stored in the same database or against technical index terms, text and/or abstracts stored in a different database.

8. A method of creating a database for a collection of scientific or technical documents, such as patents and/or technical or scientific publications and/or abstracts of these patents or publications, by:

(a) creating a hierarchical model of subjects of potential interest

(b) creating an expert technical search (ETS) for each subject-category), within the hierarchical model

(c) automatically applying the set of ETS'"'"'s against a set of electronically stored patents, patent abstracts, patent identifiers, technical index tenths, technical documents, and/or abstracts of these documents

(d) using the results of the search strategies to electronically assign each patent or technical document to one or more subject-categories within the hierarchical model

(e) Storing the set of subject-categories and unique patent identifiers or unique document identifiers in an electronic database table.

9. The method of claim 8 further comprising the step of electronically manipulating the stored subject-categories, the stored unique identifiers and the stored technical index tenths to display tabular summaries an graphical summaries of the stored data.

10. The method of claim 8 further comprising the step of using the tabular summaries and the graphical summaries to identify trends and discontinuities.

13. A method as of claim 8 where the relational database used to store, categorize, and manipulate patent abstracts can also store electronic objects such as graphic images, spreadsheets, video images and chemical structures linked to specific patent or document records.

14. A method as of claim 8 where the relational database used to store the patent abstracts and the subject-categories can display the patent abstracts, the U.S. claims, the European claims and the subject-categories on a single computer display screen by use of scrollable windows.

15. A method as of claim 8 where the database can also contain subject-specific tables of technical details such as catalyst precursors, cocatalysts, reaction conditions, reactor types, which are captured in a discrete form by scientists evaluating the underlying patents.

16. A method as of claims 8 where the database can also contain links to full-text sources of patents either in a local CD-ROM, a local LAN, a Wide-Area-LAN or on the Internet.

17. A method as of claim 8 wherein the set of technical searches are created, stored and executed on a commercial database system to identify the technology represented by each technical document or patent.

18. A method as of claim 8 wherein the set of technical searches are created and stored and executed on a local database system using documents captured from a commercial database system to identify the technology represented by each technical document or patent.

19. A method as of claim 8 wherein each technical search can automatically assign a set of technical documents or patents to one or more technical categories within a multidimensional hierarchy.

20. The method as in claim 8 further comprising the step of deriving each higher level of the multidimensional subject hierarchy from the sum of an aggregate count of unique items in each category multiplied by a weighting factors for each category in the next higher level, producing a higher (more abstract) level within the hierarchy:

space="preserve" listing-type="equation">f(x1)=sum((attribute-1*weighting-factor-1)+(attribute-2*weighting-factor-1)+ . . . )

space="preserve" listing-type="equation">f(x2)=sum((attribute-1*weighting-factor-2)+(attribute-2*weighting-factor-2)+ . . . )

space="preserve" listing-type="equation">f(x3)=sum((attribute-1*weighting-factor-3)+(attribute-2*weighting-factor-3)+ . . . ).

21. A method as in claim 8 wherein the aggregate count of unique items is determined by modeling a scientific or business discipline based on the technical content of patents or technical documents in that scientific or business discipline.

22. A method as in claim 8 wherein the weighting factors are derived by the cumulative knowledge of experts in that scientific or business discipline to reflect the impact of each technical category on the next higher level of the hierarchy.