Multimedia conceptual search system and associated search method

• US 20070130112A1
• Filed: 10/06/2006
• Published: 06/07/2007
• Est. Priority Date: 06/30/2005
• Status: Abandoned Application

• Current Assignee: IntelligenTek Corp.
• Original Assignee: IntelligenTek Corp.
• Inventors: Lin, Julius
• Application Number:
  US11/543,899
• Publication Number:
  US 20070130112A1
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: G06F 16/41 Indexing; Data structures therefor; Storage structures

Abstract

The current disclosure uses the disciplines of Ontology and Epistemology to implement a context/content-based “multimedia conceptual search and planning”, in which the formation of conceptualization is supported by embedding multimedia sensation and perception into a hybrid database. The disclosed system comprises: 1) A hybrid database model to host concept setup. 2) A graphic user interface to let user freely issue searching request in text and graphic mode. 3) A parsing engine conducting the best match between user query and dictionaries, analyzing queried images, detecting and presenting shape and chroma, extracting features/texture of an object. (4) A translation engine built for search engine and inference engine in text and graphic mode. 5) A search engine using partitioned, parallel, hashed indexes from web crawler result, conducting search in formal/natural language in text and graphic mode. 6) A logic interference engine working in text and graphic mode, and 7) A learning/feedback interface.

Claims

1. A multimedia conceptual search method comprising steps of:

providing a hybrid database including a plurality of entries, wherein each entry has at least one meaning identifier (ID), and each meaning ID is corresponding to a concept ID;

parsing a user query expression to determine at least one matching entry, which the user query expression contains, within the entries of the hybrid database;

determining at least a matching concept ID of the user query expression according to the at least one matching entry, wherein one of the at least one meaning ID of each matching entry is corresponding to the matching concept ID;

translating the user query expression to other equivalent expressions according to the entries each of which has one meaning ID corresponding to the matching concept ID; and

searching a storage media for any relevant object stored therein according to the user query expression and the other equivalent expressions.

2. The method according to claim 1, further comprising:

receiving user query data from a user interface; and

transforming the user query data into the user query expression.

3. The method according to claim 1, further comprising:

returning search results through a user interface.

4. The method according to claim 1, wherein when the user query expression is a text, the method is performed in a text mode.

5. The method according to claim 4, wherein each matching entry is one of the following:

a lexicon, a word, a term, a phrase, an idiom, a regular expression, and a sentence.

6. The method according to claim 5, wherein the meaning ID, which is corresponding to the matching concept ID, of each matching entry is determined according to grammar belongingness of the matching entry in the text.

7. The method according to claim 1, wherein when the user query expression is an image or a graphic object, the method is performed in a graphic mode.

8. The method according to claim 7, wherein each matching entry is one or one combination of the following:

a shape, a chroma, a texture, a pattern, a size, and an icon.

9. The method according to claim 8, wherein the shape is represented by a Non-Uniform Rational B-spline (NURBS) curve.

10. A multimedia conceptual search system comprising:

a hybrid database for storing a plurality of entries, wherein each entry has at least one meaning identifier (ID), and each meaning ID is corresponding to a concept ID;

a parsing engine for parsing a user query expression to determine at least one matching entry, which the user query expression contains, within the entries of the hybrid database, and for determining at least a matching concept ID of the user query expression according to the at least one matching entry, wherein one of the at least one meaning ID of each matching entry is corresponding to the matching concept ID;

a translation engine for translating the user query expression to other equivalent expressions according to the entries each of which has one meaning ID corresponding to the matching concept ID; and

a search engine for searching a storage media for any relevant object stored therein according to the user query expression and the other equivalent expressions.

11. The system according to claim 10, further comprising:

a user interface for receiving user query data and transforming the user query data into the user query expression.

12. The system according to claim 10, further comprising:

a user interface for returning search results.

13. The system according to claim 10, wherein when the user query expression is a text, the system is operated in a text mode.

14. The system according to claim 13, wherein each matching entry is one of the following:

a lexicon, a word, a term, a phrase, an idiom, a regular expression, and a sentence.

15. The system according to claim 14, wherein the parsing engine determines the meaning ID, which is corresponding to the matching concept ID, of each matching entry according to grammar belongingness of the matching entry in the text.

16. The system according to claim 10, wherein when the user query expression is an image or a graphic object, the system is operated in a graphic mode.

17. The system according to claim 16, wherein each matching entry is one or one combination of the following:

a shape, a chroma, a texture, a pattern, a size, and an icon.

18. The system according to claim 17, wherein the shape is represented by a Non-Uniform Rational B-spline (NURBS) curve.