Systems and methods for structured vocabulary search and classification

• US 6,757,692 B1
• Filed: 06/09/2000
• Issued: 06/29/2004
• Est. Priority Date: 06/09/2000
• Status: Expired due to Term

• Current Assignee: Northrop Grumman Systems Corporation (Northrop Grumman Corporation)
• Original Assignee: Northrop Grumman Corporation
• Inventors: Davis, Charles F., Ong, Kim L.
• Primary Examiner(s): Metjahic, Safet
• Assistant Examiner(s): NGUYEN, MERILYN P
• Application Number:
  US09/591,317
• Time in Patent Office: 1,481 Days
• Field of Search: 707/104, 707/5, 707/531, 707/2, 707/3, 707/4, 707/513, 707/104.1, 704/9, 704/1
• US Class Current: 707/692
• CPC Class Codes: G06F 16/353 into predefined classes
• CPC Class Codes: G06F 16/36 Creation of semantic tools, e.g. ontology or thesauri
• CPC Class Codes: Y10S 707/941 Human sciences
• CPC Class Codes: Y10S 707/99945 Object-oriented database structure processing

Abstract

Systems and methods for classifying structured vocabulary where an input is received that includes one or more terms. The one or more terms relate to an area of technology. Each one or more terms are identified and extracted from the input. The results of the extraction may be reviewed, and manually modified if appropriate. Each extracted one or more terms are classified where the classification associates a classified term to each extracted one or more terms. Each classified term is related to the area of technology. The results of the classification may be reviewed, and manually modified if appropriate. A result output is generated containing each one or more terms and the associated classified terms.

Claims

1. A method for classifying structured vocabulary comprising:

receiving input comprising at least one term, the at least one term related to an area of technology;

extracting every at least one term from the input to identify verbatim terms;

reviewing results from the extracting and manually modifying the extracted at least one term to eliminate duplicates of the extracted at least one term;

classifying each extracted at least one term, the classification associating one or more classified terms to each extracted at least one term, each of the one or more classified terms being related to the area of technology;

reviewing results from the classifying and manually modifying the classification results to eliminate one or more of the one or more classified terms; and

generating a result output containing each at least one term and the one or more associated classified terms.

2. The method according to claim 1, wherein the area of technology comprises medical.

3. The method according to claim 2, wherein the medical area of technology comprises drugs and adverse reactions to the drugs.

4. The method according to claim 1, further comprising filtering the extracted at least one term, the filtering removing the duplicates of the extracted at least one term and producing at least one unique term, the classifying being performed on the at least one unique term.

5. The method according to claim 1, wherein the at least one term comprises a medical term.

6. The method according to claim 5, wherein the medical term comprises a drug.

7. The method according to claim 5, wherein the medical term comprises an adverse reaction to a drug.

8. The method according to claim 1, wherein the input comprises narrative text.

9. The method according to claim 1, the extracting further comprising comparing the at least one term with terms in at least one database.

10. The method according to claim 9, wherein one of the at least one database comprises a thesaurus database, terms in the thesaurus database being related to the area of technology.

11. The method according to claim 10, wherein the terms in the thesaurus database are organized in a multi-tier structure.

12. The method according to claim 9, wherein at least one of the at least one database comprises a synonym database.

13. The method according to claim 9, wherein at least one of the at least one database comprises a knowledge base, terms in the knowledge base being related to the thesaurus database.

14. The method according to claim 13, wherein the terms in the knowledge base comprise at least one of acronyms, abbreviations, fillers, proximity data, suffix data, and synonymous words.

15. The method according to claim 9, wherein the comparing comprises matching each at least one term with the terms in the at least one database based on syntactical closeness, the matching using lexical matching algorithms, the matching manipulating each at least one term within a sentence.

16. The method according to claim 15, further comprising generating a goodness of fit ranking, the goodness of fit ranking being based on an amount of lexical manipulations.

17. The method according to claim 1, further comprising formatting the input into at least one input table before the extracting.

18. The method according to claim 9, further comprising updating the terms in the at least one database based on the result output.

19. The method according to claim 1, further comprising formatting the result output into at least one table.

20. The method according to claim 1, wherein the extracting every at least one term from the input and the classifying each extracted at least one term comprise extracting every at least one term from the input and classifying each extracted at least one term in an automated manner.

21. A system for classifying structured vocabulary comprising:

at least one network;

at least one client computing device, the at least one client computing device operatively connected to the at least one network;

at least one database, the at least one database operatively connected to the at least one network; and

at least one server, the at least one server operatively connected to the at least one network, at least one server receiving input from at least one client, the input comprising at least one term related to an area of technology and causing the at least one server to perform;

extracting every at least one term from the input to identify verbatim terms;

allowing review of results from the extracting and manual modification of the extracted at least one term to eliminate duplicates of the extracted at least one term;

classifying each extracted at least one term, the classification associating a classified term to each extracted at least one term, each classified term being related to the area of technology;

allowing review of results from the classifying and manual modification of the classification results; and

generating a result output containing each at least one term and the associated classified term.

22. The system according to claim 21, the extraction results review and modification being performed at one of the at least one client.

23. The system according to claim 21, the classification results review and modification being performed at one of the at least one client.

24. An article comprising a storage medium having instructions stored therein, when executed causes a computing device to perform:

receiving input comprising at least one term, the at least one term related to an area of technology;

extracting every at least one term from the input to identify verbatim terms;

reviewing results from the extracting and manually modifying the extracted at least one term to eliminate duplicates of the extracted at least one term;

classifying each extracted at least one term, the classification associating one or more classified terms to each extracted at least one term, each of the one or more classified terms being related to the area of technology;

reviewing results from the classifying and manually modifying the classification results to eliminate one or more of the one or more classified terms; and

generating a result output containing each at least one term and the one or more associated classified terms.

25. A system for permitting structured vocabulary searches, the system comprising:

an enterprise having a single point of control for all system wide decisions and controls, wherein the enterprise includes an enterprise level database for receiving textual data that contains terms to be classified;

one or more domains that are each related to a particular scientific or technological area, wherein the one or more domains are provided within the enterprises and each of the one or more domains includes one or more knowledge bases and a domain level term synonym table;

one or more work packages provided within each of the one or more domains, wherein each of the one or more work packages defines a particular organization structure for effecting an evaluation, wherein each of the one or more work packages comprises;

a text extractor for evaluating narrative text, for extracting verbatim terms from the narrative text and for assigning a label to each of the verbatim terms;

a classifier for evaluating each of the verbatim terms extracted by the text extractor and deriving preferred matching and classification terms based upon one of the one or more knowledge bases or the domain level term synonym table; and

a reviewer for reviewing the classification of verbatim terms that have been classified by the classifier, for reviewing the verbatim terms extracted by the text extractor and for making corrections,

wherein the enterprise includes an enterprise agent for storing the received textual data in an appropriate table based on a particular work package associated with the textual data and for later sending the received textual data to the particular work package.

26. The system according to claim 25, further comprising an authorized director for creating and maintaining each of the one or more domains.

27. The system according to claim 25, further comprising an authorized administrator for creating and maintaining each of the one or more work packages.

28. The system according to claim 25, further comprising an authorized work manager for managing each of the one or more work packages.

29. The system according to claim 25, wherein the enterprise agent sends the received textual data to the particular work package via an Open Database Connectivity connection.

30. A system for permitting structured vocabulary searches, the system comprising:

an enterprise having a single point of control for all system wide decisions and controls, wherein the enterprise includes an enterprise level database for receiving textual data that contains terms to be classified;

one or more domains that are each related to a particular scientific or technological area, wherein the one or more domains are provided within the enterprises and each of the one or more domains includes one or more knowledge bases and a domain level term synonym table;

one or more work packages provided within each of the one or more domains, wherein each of the one or more work packages defines a particular organization structure for effecting an evaluation, wherein each of the one or more work packages comprises;

a text extractor for evaluating narrative text, for extracting verbatim terms from the narrative text and for assigning a label to each of the verbatim terms;

a classifier for evaluating each of the verbatim terms extracted by the text extractor and deriving preferred matching and classification terms based upon one of the one or more knowledge bases or the domain level term synonym table; and

a reviewer for reviewing the classification of verbatim terms that have been classified by the classifier, for reviewing the verbatim terms extracted by the text extractor and for making corrections,

wherein the enterprise includes an enterprise agent for storing the received textual data in an appropriate table based on a particular work package associated with the textual data and for later sending the received textual data to the particular work package,

wherein;

a text extractor identification is recorded to permit audit of decisions made by the text extractor;

a classifier identification is recorded to permit audit of decisions made by the classifier; and

a reviewer identification is recorded to permit audit of decisions made by the reviewer.

31. A method for searching and classifying structured vocabulary, the method comprising:

extracting received input data to obtain verbatim terms, wherein the extracting of the received input data includes;

formatting the input data into a work table of terms;

comparing the work table of terms with previous input terms and customer term synonym lists; and

defining terms in the work table of terms that are identifiable with the previous input terms or the synonym lists as the verbatim terms;

manually reviewing the verbatim terms to eliminate duplicate verbatim terms;

comparing the verbatim terms to concepts related to scientific or technological areas to determine if the verbatim terms are associated with any of the specific scientific or technological areas and accordingly classifying the verbatim terms in one or more of the scientific or technological areas if such an association is determined;

manually reviewing the classified verbatim terms; and

generating a result output that includes the classified verbatim terms, wherein the result output is in a format appropriate for storage in a customer database.

32. The method of claim 31, wherein extracting of the received input data is performed by a matching algorithm that utilizes lexical characteristics augmented with knowledge of a specific area of science or technology.