Computer method for searching document and recognizing concept with controlled tolerance

• US 20140304291A9
• Filed: 03/05/2012
• Published: 10/09/2014
• Est. Priority Date: 05/24/2006
• Status: Abandoned Application

• Current Assignee: Sizhe Tan
• Original Assignee: Sizhe Tan
• Inventors: Tan, Sizhe
• Application Number:
  US13/385,735
• Publication Number:
  US 20130232166A1
• Time in Patent Office: Days
• US Class Current: 707/769
• CPC Class Codes: G06F 16/31 Indexing; Data structures therefor; Storage structures
• CPC Class Codes: G06F 16/583 using metadata automatically derived from the content

Abstract

Documents are searched by a target document. The target document is tokenized into buta strings. The buta strings are decomposed into buta attribute values. A target buta attribute value is selected. A tolerance is given to the target buta attribute value. A buta attribute range is determined from the given tolerance. Buta attribute value suggestions are lookup results within the buta attribute range in dictionary of index II, which relates a buta attribute value to buta strings. Alternative buta strings are searched using the buta attribute value suggestions in dictionary of index II. Finally, documents can be searched using the alternative buta strings in dictionary of index I, which relates a buta string to documents.

Claims

1. A computer method for searching image documents comprising the computer executed steps of:

tokenizing an query into image pattern tokens

representing said image pattern tokens with buta strings,

decomposing said buta strings into buta attribute values,

selecting a target buta attribute value from said buta attribute values,

giving a tolerance to said target buta attribute value,

determining a buta attribute range using said given tolerance,

searching buta attribute value suggestions in a dictionary of index II within said buta attribute range, wherein said dictionary of index II relates a buta attribute value to buta strings,

obtaining alternative buta strings comprising OR operations among said buta attribute value suggestions,

searching image documents comprising OR operations among said alternative buta strings in a dictionary of index I, wherein said dictionary of index I relates a buta string to image documents.

2. The computer method of claim 1 further comprising:

searching image documents comprising AND operations among said buta strings in said dictionary of index I.

3. The computer method of claim 1 further comprising:

providing a plurality of image documents,

tokenizing each of said plurality of image documents into image pattern tokens,

collecting said image pattern tokens in a master collection of image pattern tokens,

transforming said master collection of image pattern tokens into said dictionary of index I by representing said image pattern tokens with buta strings,

constructing said dictionary of index II by decomposing said buta strings into buta attribute values.

4. The computer method of claim 3 wherein said provided image documents are from a data source.

5. The computer method of claim 3 wherein said provided image documents are from the Internet.

6. A computer method for searching documents comprising the computer executed steps of:

tokenizing a target document into buta strings,

decomposing said buta strings into buta attribute values,

selecting a target buta attribute value from said buta attribute values,

giving a tolerance to said target buta attribute value,

determining a buta attribute range using said given tolerance,

searching buta attribute value suggestions in a dictionary of index II within said buta attribute range, wherein said dictionary of index II relates a buta attribute value to buta strings,

obtaining alternative buta strings comprising OR operations among said buta attribute value suggestions,

searching documents comprising OR operations among said alternative buta strings in a dictionary of index I, wherein said dictionary of index I relates a buta string to documents.

7. The computer method of claim 6 further comprising:

searching documents comprising AND operations among said buta strings in said dictionary of index I.

8. The computer method of claim 6 further comprising:

providing a plurality of documents,

tokenizing each of said plurality of documents into buta strings,

collecting said buta strings into said dictionary of index I,

constructing said dictionary of index II by decomposing said buta strings into buta attribute values.

9. The computer method of claim 8 wherein said provided documents are from a data source.

10. The computer method of claim 8 wherein said provided image documents are from the Internet.

11. A computer method for recognizing concepts comprising the computer executed steps of:

tokenizing a target concept into buta strings,

decomposing said buta strings into buta attribute values,

selecting a target buta attribute value from said buta attribute values,

giving a tolerance to said target buta attribute value,

determining a buta attribute range using said given tolerance,

searching buta attribute value suggestions in a dictionary of index II within said buta attribute range, wherein said dictionary of index II relates a buta attribute value to buta strings,

obtaining alternative buta strings comprising OR operations among said buta attribute value suggestions,

recognizing concepts comprising OR operations among said alternative buta strings in a dictionary of index I, wherein said dictionary of index I relates a buta string to concepts.

12. The computer method of claim 11 further comprising:

recognizing concepts comprising AND operations among said buta strings in said dictionary of index I.

13. The computer method of claim 11 further comprising:

providing a plurality of concepts,

tokenizing each of said plurality of concept into buta strings,

collecting said buta strings into said dictionary of index I,

constructing said dictionary of index II by decomposing said buta strings into buta attribute values.