Information searching system and method thereof

• US 20060195433A1
• Filed: 04/03/2006
• Published: 08/31/2006
• Est. Priority Date: 07/06/2000
• Status: Abandoned Application

• Current Assignee: Si Han Kim
• Original Assignee: Si Han Kim
• Inventors: Kim, Si Han
• Application Number:
  US11/397,964
• Publication Number:
  US 20060195433A1
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: G06F 16/3329 Natural language query formulation or dialogue systems
• CPC Class Codes: G06F 16/3338 Query expansion
• CPC Class Codes: G06F 16/3344 using natural language analysis
• CPC Class Codes: G06F 16/3347 using vector based model

Abstract

In an information searching method, it is first determined if an input retrieval word is a sentence composed of a plurality of words. A functioning code is assigned to each word of the sentence according to its function in the sentence, and then the words are coded in basic words. Then, a database in which information is composed of a sentence having a plurality of words each of which is assigned with a functioning code and coded in a basic word is retrieved on the basis of the coded retrieval word to search information having function codes and word codes identical to those of each word of the retrieval word.

Claims

1. An information searching system comprising:

an input part for inputting retrieval words representing information;

a database for storing word codes formed by coding words representing the information, each of the word codes being assigned with a functioning code representing its function in the information; and

a processor for coding the retrieval words into basic word codes each having a functioning code and retrieving the database on the basis of the basic word codes to search the information having identical functioning and word codes to the basic word codes,

wherein the search of information having the identical or most similar word codes is performed separately from the search of information having the identical or most similar functioning codes.

2. An information searching system of claim 1 wherein when the retrieval command includes a phrase, each of the words of the command is assigned with the functioning code such that its functions in the command and the phrase can be discriminated from each other.

3. An information searching system of claim 1 wherein when the retrieval command is formed of at least two sentences, each of the words of the sentences is assigned with the functioning code such that the sentences can be discriminated from each other.

4. An information searching system of claim 1 wherein when there is no information having the identical functioning and word codes, the processor searches information which is identical to the functioning code, and is most similar to the basic word code.

5. A method for searching information, comprising the steps of:

determining if input retrieval command is formed of plural words;

coding each of the words into a basic word codes with a functioning code; and

retrieving a database storing word codes formed by coding words representing the information on the basis of the basic word codes to search the information having identical functioning and word codes to the basic word codes.

6. A method for searching information of claim 5 wherein the step of retrieving further comprising the steps of:

selecting information which is most identical to the functioning and word codes of the words of the retrieval command except for a subject word of the retrieval command; and

searching information having a word code modified by the selected information and most identical to the subject word.

7. A method for searching information of claim 5 wherein when there are more than two words having an identical functioning code in the words of the retrieval command, the words having the identical functioning code are grouped and information having the identical functioning code and the most identical word codes is searched.

8. A method for searching information of claim 5 wherein the step of retrieving further comprises the step of searching information which is identical to a subject word code of the retrieval command and most identical to the rest word codes of the retrieval command.

9. A method for searching information comprising the steps of:

storing word codes of words representing information in a database;

coding words of a retrieval command into basic word codes according to a predetermined rule; and

searching information which is most identical to the basic word codes by retrieving the database,

wherein the word code of the retrieval command is extended to more than two word codes.

10. A method of claim 9 wherein when there is a word code of a lower level of the retrieval command not including a retrieval word code, the searching is performing on the basis of a lower level word code not including the retrieval word code.

11. A method of claim 9 wherein when the word of the retrieval command is a basic word, the word is coded as a new code formed by other basic words describing the word of the retrieval command, and the searching is performed on the basis of the new code.

12. A method of claim 9 wherein when coding the words representing the information and the words of the retrieval command, each of the words is coded including a property of the word as a constituting word code.

13. A method of claim 9 wherein when there is a word, which is not coded, among the words of the retrieval command, information including the word which is not coded is searched.

18. A method of claim 9 wherein if there is a multiple meaning word in the words of the retrieval command or the information to be retrieved, word code sets representing basic word codes constituting word code of the multiple meaning word as other basic word codes are formed, and the word code sets are compared with normal word codes.

14. A method for searching information comprising the steps of:

storing word codes of words representing information in a database;

coding words of a retrieval command into basic word codes according to a predetermined rule; and

searching information which is most identical to the basic word codes by retrieving the database,

wherein the information to be retrieved is represented as a vector value in a vector space having axe formed by basic word;

an angle α

between a basic vector and an information vector to be retrieved, is calculated and

an indexing database of the information according to the calculated angle is made.

15. A method of claim 14 wherein the words of the retrieval command are converted into a vector value, an angle Sα

between the basic vector and the retrieval word vector is calculated, and the information is searched on the basis of the calculated angle Sα

through the indexing database.

16. A method of claim 14 wherein a vector value of the retrieval word according to the functioning code is calculated in the vector space, an angle between the vector value and the basic vector is calculated, and the information is searched considering the functioning code.

17. A method of claim 14 wherein a vector value of the retrieval word according to the functioning code is calculated in the vector space, an angle between the vector value and the basic vector is calculated, and the information is searched without considering the functioning code.

27. A method of claim 14 wherein if there is a multiple meaning word in the words of the retrieval command or the information to be retrieved, word code sets representing basic word codes constituting word code of the multiple meaning word as other basic word codes are formed, and the word code sets are compared with normal word codes.

19. (canceled)

20. (canceled)

21. (canceled)

22. (canceled)

23. (canceled)

24. (canceled)

25. (canceled)

26. (canceled)