Systems and methods for generation of searchable structures respective of multimedia data content
First Claim
1. A method for generating concept structures from signature reduced clusters (SRCs), comprising:
- retrieving at least one current SRC including a cluster of signatures respective of a plurality of multimedia data elements (MMDEs);
generating at least one metadata for each signature of the cluster of signatures;
identifying a number of repetitions of a metadata of the at least one generated metadata;
determining whether the number of repetitions of the metadata of the at least one generated metadata, within the at least one current retrieved SRC, exceeds a predefined repetition threshold;
upon determining that the number of repetitions of the metadata exceeds the predefined repetition threshold, identifying the at least one metadata as representative of the SRC;
comparing the representative metadata of the current retrieved SRC to metadata that is representative of at least one previously generated SRC to determine a metadata match; and
upon determining that the metadata match, identifying the current retrieved SRC and the matching previously generated SRC as a concept structure, wherein the concept structure is represented by the representative matching metadata from both the current retrieved SRC and the at least one previously generated SRC.
3 Assignments
0 Petitions
Accused Products
Abstract
Methods and systems for generating concept structures from signature reduced clusters (SRCs) are provided. The method includes retrieving at least one SRC including a cluster of signatures respective of a plurality of multimedia data elements (MMDEs); generating at least one metadata for each signature of the cluster of signatures; identifying a number of repetitions of a metadata of the at least one generated metadata; and determining whether the number of repetitions of the metadata exceeds a predefined repetition threshold; upon determining that the number of repetitions of the metadata exceeds the predefined repetition threshold, identifying the metadata as representative of the SRC; comparing the representative metadata to metadata that is representative of at least one previously generated SRC to determine a metadata match; and upon determining the metadata match, identifying the retrieved SRC and the matching previously generated SRC as a concept structure.
-
Citations
18 Claims
-
1. A method for generating concept structures from signature reduced clusters (SRCs), comprising:
-
retrieving at least one current SRC including a cluster of signatures respective of a plurality of multimedia data elements (MMDEs); generating at least one metadata for each signature of the cluster of signatures; identifying a number of repetitions of a metadata of the at least one generated metadata; determining whether the number of repetitions of the metadata of the at least one generated metadata, within the at least one current retrieved SRC, exceeds a predefined repetition threshold; upon determining that the number of repetitions of the metadata exceeds the predefined repetition threshold, identifying the at least one metadata as representative of the SRC; comparing the representative metadata of the current retrieved SRC to metadata that is representative of at least one previously generated SRC to determine a metadata match; and upon determining that the metadata match, identifying the current retrieved SRC and the matching previously generated SRC as a concept structure, wherein the concept structure is represented by the representative matching metadata from both the current retrieved SRC and the at least one previously generated SRC. - View Dependent Claims (2, 3, 4, 5, 6)
-
-
7. A method for generating concept structures from signature reduced clusters (SRCs), comprising:
-
retrieving at least one current SRC comprising a cluster of signatures respective of a plurality of multimedia data elements (MMDEs); generating a plurality of metadata for each signature of the cluster of signatures; identifying a number of repetitions of each metadata, within the at least one current retrieved SRC, of the plurality of metadata; determining whether the number of repetitions of each metadata of the plurality of metadata exceeds a predefined repetition threshold; upon determining that the number of repetitions of a metadata of the plurality of metadata exceeds the predefined repetition threshold, identifying the metadata as representative of the SRC; determining an overlap level between all identified representative metadata of the current retrieved SRC and all metadata that is representative of at least one previously generated SRC; determining whether the overlap level exceeds a predetermined overlap threshold; and upon determining that the overlap level exceeds a predetermined overlap threshold, identifying the retrieved SRC and the previously generated SRC as a concept structure, wherein the concept structure is represented by the representative matching metadata from both the current retrieved SRC and the at least one previously generated SRC. - View Dependent Claims (8, 9, 10)
-
-
11. A system for generating concept structures from signature reduced clusters (SRCs), comprising:
-
a processor; and a memory, the memory containing instructions that, when executed by the processor, configure the system to; retrieve at least one current SRC comprising a cluster of signatures respective of a plurality of multimedia data elements (MMDEs); generate at least one metadata for each signature of the cluster of signatures; identify a number of repetitions of a metadata, within the at least one SRC, of the at least one generated metadata; determine whether the number of repetitions of the metadata exceeds a predefined repetition threshold; upon determining that the number of repetitions of the metadata exceeds the predefined repetition threshold, identify the metadata as representative of the SRC; compare the representative metadata of the at least one current retrieved SRC to metadata that is representative of at least one previously generated SRC to determine a metadata match; and upon determining that the metadata match, identify the retrieved SRC and the matching previously generated SRC as a concept structure, wherein the concept structure is represented by the representative matching metadata from both the current retrieved SRC and the at least one previously generated SRC. - View Dependent Claims (12, 13, 14, 15)
-
-
16. A system for generating concept structures from signature reduced clusters (SRCs), comprising:
- a processor; and
a memory, the memory containing instructions that, when executed by the processor, configure the system to; retrieve at least one current SRC comprising a cluster of signatures respective of a plurality of multimedia data dements (MMDEs); generate a plurality of metadata for each signature of the duster of signatures;
identify a number of repetitions of each metadata, within the at least one current retrieved SRC, of the plurality of metadata;determine whether the number of repetitions of each metadata of the plurality of metadata exceeds a predefined repetition threshold; upon determining that the number of repetitions of a metadata of the plurality of metadata exceeds the predefined repetition threshold, identify the metadata as representative of the current retrieved SRC; determine an overlap level between all identified representative metadata of the current retrieved SRC and all metadata that is representative of at least one previously generated SRC; determine whether the overlap level exceeds a predetermined threshold; and
upon determining that the overlap level exceeds a predetermined threshold, identify the retrieved SRC and the previously generated SRC as a concept structure, wherein the concept structure is represented by the representative matching metadata from both the current retrieved SRC and the at least one previously generated SRC. - View Dependent Claims (17, 18)
- a processor; and
Specification