Method for natural language data processing using morphological and part-of-speech information

• US 5,331,556 A
• Filed: 06/28/1993
• Issued: 07/19/1994
• Est. Priority Date: 06/28/1993
• Status: Expired due to Fees

• Current Assignee: Lockheed Martin Corporation (Martin Marietta Corporation)
• Original Assignee: General Electric Company
• Inventors: Zernik, Uri, Black, James E. Jr.
• Primary Examiner(s): MCELHENY JR, DONALD E
• Application Number:
  US08/082,710
• Time in Patent Office: 386 Days
• Field of Search: 364/419.01, 364/419.02, 364/419.04, 364/419.05, 364/419.08, 364/419.13
• US Class Current: 704/9
• CPC Class Codes: G06F 40/253 Grammatical analysis; Style critique

Abstract

An enhancement and retrieval method for natural language data using a computer is disclosed. The method includes executing linguistic analysis upon a text corpus file to derive morphological, part-of-speech information as well as lexical variants corresponding to respective corpus words. The derived linguistic information is then used to construct an enhanced text corpus file. A query text file is linguistically analyzed to construct a plurality of trigger token morphemes which are then used to construct a search mask stream which is correlated with the enhanced text corpus file. A match between the search mask stream and the enhanced corpus file allows a user to retrieve selected portions of the enhanced text corpus.

Claims

1. A method for constructing an enhanced text corpus file using a computer and comprising the steps of:

providing a text corpus file to said computer, said text corpus file comprising respective electrical signals representative of a predetermined natural language data;

processing said electrical signals to parse said text corpus file into a plurality of sentences each constituted of a respective stream of corpus words;

executing linguistic analysis upon each said stream of corpus words to derive respective part-of-speech information and morphological roots corresponding to respective ones of said corpus words; and

generating an enhanced text corpus file using said derived morphological roots and said derived part-of-speech information.

2. A method in accordance with claim 1 wherein the step of executing said linguistic analysis comprises:

executing morphological analysis upon each said stream of corpus words to derive said morphological roots; and

executing syntactic analysis upon each said stream of corpus words to derive said part-of-speech information.

3. A method in accordance with claim 2 wherein the step of executing said syntactic analysis further includes tagging said part-of-speech information with respective ones of said corpus words.

4. A method in accordance with claim 3 wherein the step of executing said linguistic analysis further includes executing lexical analysis to acquire predetermined lexical entries corresponding to respective ones of said corpus words.

5. A method in accordance with claim 4 wherein the step of executing said linguistic analysis further includes executing semantic analysis upon each said stream of corpus words to generate lexical variations corresponding to respective ones of said corpus words.

6. A method in accordance with claim 5 wherein said enhanced text corpus file comprises a plurality of enhanced sentences each constituted of a respective stream of enhanced words.

7. A method in accordance with claim 6 wherein each of said enhanced words includes at least the derived morphological root and the part-of-speech information corresponding to said enhanced word.

8. A method in accordance with claim 7 wherein each of said enhanced words further includes a corresponding affix.

9. A method in accordance with claim 8 wherein said corresponding affix comprises a prefix and a suffix.

10. A method for constructing trigger token morphemes using a computer and comprising the steps of:

providing a query text file to said computer, said query text file comprising respective electrical signals representative of predetermined inquiry data;

processing the electrical signals representative of said predetermined inquiry data to parse said query text file into respective query items each constituted of a respective stream of query words;

executing morphological analysis upon each said stream of query words to derive respective morphological roots corresponding to respective ones of said query words;

executing semantic analysis upon each said stream of query words to generate respective lexical variants corresponding to respective ones of said query words, and

generating a plurality of trigger token morphemes corresponding to respective ones of said query items, said plurality of trigger token morphemes using said derived morphological roots and said derived lexical variants corresponding to respective ones of said query words.

11. A method according to claim 10 wherein the step of providing said query text file further comprises providing part-of-speech and collocation information associated with respective ones of said query words.

12. A method according to claim 11 further comprising the step of executing domain specific synonym analysis upon each said stream of query words to derive predetermined synonyms corresponding to respective ones of said query words.

13. A method for retrieving selected portions from an enhanced text corpus file using a computer and comprising the steps of:

generating a search mask stream based upon a plurality of predetermined trigger token morphemes;

scanning said enhanced text corpus file; and

correlating said search mask stream with respect to said enhanced text corpus file for retrieving a selected portion of said enhanced text corpus file based upon a match between said search mask stream and said enhanced text corpus file.

14. An enhancement and retrieval method for natural language data using a computer and comprising the steps of:

providing a text corpus file to said computer, said text corpus file comprising respective electrical signals representative of said natural language data;

processing said electrical signals to parse said text corpus file into a plurality of sentences each constituted of a respective stream of corpus words;

executing linguistic analysis upon each said stream of corpus words to derive respective part-of-speech information and morphological roots corresponding to respective ones of said corpus words; and

generating an enhanced text corpus file using said derived morphological roots and said derived part-of-speech information.

15. A method in accordance with claim 14 wherein the step of executing said linguistic analysis comprises:

executing morphological analysis upon each said stream of corpus words to derive said morphological roots; and

executing syntactic analysis upon each said stream of corpus words to derive said part-of-speech information.

16. A method in accordance with claim 15 wherein the step of executing said syntactic analysis further includes tagging said part-of-speech information with respective ones of said corpus words.

17. A method in accordance with claim 16 wherein the step of executing said linguistic analysis further includes executing lexical analysis to acquire predetermined lexical entries corresponding to respective ones of said corpus words.

18. A method in accordance with claim 17 wherein the step of executing said linguistic analysis further includes executing semantic analysis upon each said stream of corpus words to generate lexical variations corresponding to respective ones of said corpus words.

19. A method in accordance with claim 18 wherein said enhanced text corpus file comprises a plurality of enhanced sentences each constituted of a respective stream of enhanced words.

20. A method in accordance with claim 19 wherein each of said enhanced words includes at least the derived morphological root and the part-of-speech information corresponding to said enhanced word.

21. A method in accordance with claim 20 wherein each of said enhanced words further includes a corresponding affix.

22. A method in accordance with claim 21 wherein said corresponding affix comprises a prefix and a suffix.

23. A method in accordance with claim 22 and further comprising the steps of:

providing a query text file to said computer, said query text file comprising respective electrical signals representative of predetermined inquiry data;

processing the electrical signals representative of said predetermined inquiry data to parse said query text file into respective query items each constituted of a respective stream of query words;

executing morphological analysis upon each said stream of query words to derive respective morphological roots corresponding to respective ones of said query words;

executing semantic analysis upon each said stream of query words to generate respective lexical variants corresponding to respective ones of said query words, and

generating a plurality of trigger token morphemes corresponding to respective ones of said query items, said plurality of trigger token morphemes using said derived morphological roots and said derived lexical variants corresponding to respective ones of said query words.

24. A method according to claim 23 wherein the step of providing said query text file further comprises providing part-of-speech and collocation information associated with respective ones of said query words.

25. A method according to claim 24 further comprising the step of executing domain specific synonym analysis upon each said stream of query words to derive predetermined synonyms corresponding to respective ones of said query words.

26. A method according to claim 25 further comprising the steps of:

generating a search mask stream based upon said plurality of trigger token morphemes;

scanning said enhanced text corpus file; and

correlating said search mask stream with respect to said enhanced text corpus file for retrieving a selected portion of said enhanced text corpus file based upon a match between said search mask stream and said enhanced text corpus file.