System and methods for generation of a concept based database

• US 9,672,217 B2
• Filed: 03/10/2015
• Issued: 06/06/2017
• Est. Priority Date: 10/26/2005
• Status: Active Grant

• Current Assignee: Cortica Ltd.
• Original Assignee: Cortica Ltd.
• Inventors: Raichelgauz, Igal, Odinaev, Karina, Zeevi, Yehoshua Y.
• Primary Examiner(s): LE, UYEN T
• Application Number:
  US14/643,694
• Publication Number:
  US 20150186405A1
• Time in Patent Office: 819 Days
• Field of Search: None
• CPC Class Codes: G06F 16/1748 De-duplication implemented within the file system, e.g. based on file segments de-duplication techniques in storage systems for the management of data blocks G06F3/0641
• CPC Class Codes: G06F 16/2228 Indexing structures
• CPC Class Codes: G06F 16/285 Clustering or classification
• CPC Class Codes: G06F 16/41 Indexing; Data structures therefor; Storage structures
• CPC Class Codes: G06F 16/43 Querying
• CPC Class Codes: Y10S 707/99943 Generating database or data structure, e.g. via user interface
• CPC Class Codes: Y10S 707/99948 Application of database or data structure, e.g. distributed, multimedia, or image

Abstract

A system and method for generating a concept database based on at least two multimedia data elements (MMDEs). The method includes: generating at least two items from a received MMDE of the at least two MMDEs; determining the items that are of interest for signature generation; generating at least one signature responsive to at least one item of interest of the received MMDE; clustering at least two signatures received from the signature generator responsive of the plurality of MMDEs; reducing the number of signatures in each cluster to a create a signature reduced cluster (SRC) of the cluster; associating metadata with the SRC to a concept structure including at least two SRCs and their associated metadata; and generating at least one index for mapping the received MMDE to at least one concept structure, wherein the concept database includes concept structures and the generated indices for the at least two MMDEs.

Claims

1. A system for generating a concept database based on a plurality of multimedia data elements (MMDEs), comprising:

an attention processor (AP) configured to generate a plurality of items from a received MMDE of the plurality of MMDEs and to determined which of the generated items are of interest for signature generation, wherein an item is of interest for signature generation if an entropy of the item is above a predefined threshold;

a signature generator (SG) configured to generate at least one signature responsive to at least one item of interest of the received MMDE of the plurality of MMDEs;

a clustering processor (CP) configured to cluster a plurality of signatures received from the signature generator responsive of the plurality of MMDEs and to create a signature reduced cluster (SRC);

a concept generator (CG) configured to associate metadata with the SRC and to form a concept structure comprised of a plurality of previously generated SRCs, the created SRC, and their associated metadata, wherein the CG is further configured to;

generate metadata for each of the created SRC,

compare the generated metadata to metadata associated with each SRC of the plurality of previously generated SRCs to determine if there is a match, and

associate the created SRC with the matching SRCs of the plurality of previously generated SRCs upon determining that there is a match; and

an index generator (IG) configured to generate at least one index for mapping the received MMDE to the at least one concept structure formed by the CG, wherein the at least concept structures and the at least one generated index are stored in the concept database.

2. The system of claim 1, wherein the CG is further configured to:

create a list of the generated metadata.

3. The system of claim 2, wherein the CG is further configured to:

remove duplicated metadata of the list of the generated metadata.

4. The system of claim 2, wherein the CG is further configured to:

determine a number of repetitions of each metadata of the list of the generated metadata.

5. The system of claim 4, wherein the CG is further configured to:

compare the number of repetitions of a compared metadata of the list of the generated metadata to a threshold; and

upon determining that the number of repetitions is below a threshold, remove all instances of the compared metadata from the list of the generated metadata.

6. The system of claim 4, wherein the CG is further configured to:

determine a strength of each metadata of the list of the generated metadata based on each respective number of repetitions.

7. The system of claim 1, wherein the IG is further configured to:

index the plurality of MMDEs based on the concept structure.

8. The system of claim 1, wherein the AP is a patch AP (PAP) that is configured to extract items in the form of patches from the received MMDE.

9. The system of claim 8, wherein the PAP is further configured to:

determine which patches to provide for the signature generation based on at least one of;

an entropy level of a patch, corners identified in a patch, and borders identified in a patch.

10. The system of claim 8, wherein the associated metadata is at least one of:

a plurality of patches created by the PAP from each MMDE, the at least one signature generated by the SG to each MMDE, and the concept matched for each MMDE based on the signatures generated by the CP.

11. A method for generating a concept database based on a plurality of multimedia data elements (MMDEs) comprising:

generating a plurality of items from a received MMDE of the plurality of MMDEs;

determining the items that are of interest for signature generation, wherein an item is of interest for signature generation if an entropy of the item is above a predefined threshold;

generating at least one signature responsive to at least one item of interest of the received MMDE of the plurality of MMDEs;

clustering a plurality of signatures received from the signature generator responsive of the plurality of MMDEs;

reducing the number of signatures in each cluster to create a signature reduced cluster (SRC) of the cluster;

associating metadata with the SRC and forming a concept structure comprised of a plurality of previously generated SRCs, the created SRC, and their associated metadata, wherein the associating further comprises;

generating metadata for each of the created SRC,

comparing the generated metadata to metadata associated with each SRC of the plurality of previously generated SRCs to determine if there is a match, and

upon determining that there is a match, associating the created SRC with the matching SRCs of the plurality of previously generated SRCs;

generating at least one index for mapping the received MMDE to the at least one concept structure; and

storing the at least one concept structure and the at least one generated index in the concept database.

12. The method of claim 11, further comprising:

creating a list of the generated metadata.

13. The method of claim 12, further comprising:

removing duplicated metadata of the list of the generated metadata.

14. The method of claim 12, further comprising:

determining a number of repetitions of each metadata of the list of the generated metadata.

15. The method of claim 14, further comprising:

comparing the number of repetitions of a compared metadata of the list of the generated metadata to a threshold; and

upon determining that the number of repetitions is below a threshold, removing all instances of the compared metadata from the list of the generated metadata.

16. The method of claim 14, further comprising:

determining a strength of each metadata of the list of the generated metadata based on each respective number of repetitions.

17. The method of claim 11, further comprising:

indexing the plurality of MMDEs based on the concept structure.

18. The method of claim 11, wherein the AP is a patch AP (PAP) that is configured to extract items in the form of patches from the received MMDE.

19. The method of claim 18, wherein the PAP is further configured to:

determine which patches to provide for the signature generation based on at least one of;

an entropy level of a patch, corners identified in a patch, and borders identified in a patch.

20. The method of claim 18, wherein the associated metadata is at least one of:

a plurality of patches created by the PAP from each MMDE, the at least one signature generated by the SG to each MMDE, and the concept matched for each MMDE based on the signatures generated by the CP.

21. A non-transitory computer readable medium having stored thereon instructions for causing one or more processing units to execute the method according to claim 11.