DETAILED ACTION
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Information Disclosure Statement
IDS submitted 7/21/2025 has been considered by the examiner.
Claim Rejections - 35 USC § 103
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis (i.e., changing from AIA to pre-AIA ) for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
Claim 1-3, 6, 8, 10-12, 15, 17 and 19-20 are rejected under 35 U.S.C. 103 as being unpatentable over Yuan, Xin et al (PGPUB Document No. 20230260164), hereafter referred as to “Yuan”, in view of Liu, Shu et al (PGPUB Document No. 20210350183), hereafter, referred to as “Liu”, in view of Sahu, Tezan et al (PGPUB Document No. 20240411824), hereafter, referred to as “Sahu”.
Claim 1, Yuan teaches A content search method performed by a computer device, comprising: obtaining search information and a media resource, the media resource comprising a plurality of pieces of media content(Yuan, para 0004 discloses searching media/image based on extracted features “Embodiments of the inventive concept involve doing an image search based on the text prompt, and then using one or more retrieved images along with the text to generate a new image …….”); extracting a text feature from the search information (Yuan, para 0080 discloses extracting text features from search or query “text phrase 600 is input to a text encoder to generate text features 610. As an example, text phrase 600 states “a couple of baseball players on a field”, which is a query from user 100”) and a content feature from each of the plurality pieces of media content(Yuan, para 0079 discloses extraction content features from images “image attention network 635 is applied to the search image to obtain image attention features”);
performing semantic recognition on the mapped features based on the text feature, to determine semantic types corresponding to the mapped features(yuan, para 0103 discloses image/content features are compared for recognition based semantic information or feature in the query text “the target image includes substantially similar semantic features to semantic information as indicated in the text phrase”);
But Yuan does not explicitly teach transforming the plurality of content features to multiple mapped features, wherein a distance between a pair of mapped features represents semantic relevance between the pair of mapped features; grouping the mapped features corresponding to the same semantic type into a same combination; determining target mapped features meeting a relevance condition from different combinations based on the distances between mapped features in the different combinations; and determining search results for the search information from the media resource according to the target mapped features.
However, in the same field of endeavor of content feature analysis Liu teaches transforming the plurality of content features to multiple mapped features(Liu, para 0058 discloses transforming obtained feature through mapping “The computer device can map an obtained semantic feature to the instance feature space through the mapping relationship between the semantic feature space and the instance feature space, to obtain a transformed feature……”), wherein a distance between a pair of mapped features represents semantic relevance between the pair of mapped features(Liu, para 0079 discloses further discloses distances between features are Euclidean distance (similarity distance) “The feature distances are vector distances between the semantic-fused instance feature vector of each point and semantic-fused instance feature vectors of other points in the point cloud except each point. For example, the vector distances herein may be Euclidean distances or other distances”);
grouping the mapped features corresponding to the same semantic type into a same combination(Liu, para 0089 discloses grouping/clustering features by according to their closeness in semantic distance “the computer device may perform clustering with each point in the point cloud as a central point according to the feature distances between the semantic-fused instance features of the points in the point cloud, and find a plurality of adjacent points (including the central point itself) adjacent to each point. The plurality of adjacent points form a set of points, and a set of points may be considered as a local point cloud”);
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate the feature of clustering content features based on their similarity of Liu into extraction of features of Liu to produce an expected result of mapping two sets of features based on their similarity. The modification would be obvious because one of ordinary skill in the art would be motivated to accurately categorize content features based on their semantic similarity distance (Liu, para 0089) .
But Yuan and Liu don’t explicitly teach determining target mapped features meeting a relevance condition from different combinations based on the distances between mapped features in the different combinations;
and determining search results for the search information from the media resource according to the target mapped features.
However, in the same field of endeavor of content feature analysis Sahu teaches determining target mapped features meeting a relevance condition from different combinations based on the distances between mapped features in the different combinations(Sahu, para 0088 discloses mapping obtained features with target features based on threshold distance of matching “the query reformulation model 530 generates a feature vector for the provided autosuggest query and determines if the feature vector maps within a threshold distance to a feature vector of a known autosuggest query within the learned vector space”); and determining search results for the search information from the media resource according to the target mapped features(Sahu, element 310 of Fig. 3 and para 0048 disclose obtaining result based on target or newly mapped features (auto suggested query) “In the first path corresponding to search results 310, the flow 300 includes act 312 of an autosuggest query being selected. For example, a user selects one of the autosuggest queries. In response, the autosuggest query system updates the first user interface to display the search results corresponding to the selection”).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate the feature of mapping based on similarity distance of Sahu into mapping of content features of Yuan and Liu to produce an expected result of mapping two sets of features based on their similarity. The modification would be obvious because one of ordinary skill in the art would be motivated to improve search computational efficiency and accuracy (Sahu, abstract) .
Regarding claim 2, Yuan, Liu and Sahu teach all the limitations of claim 1 and Yuan further teaches wherein the extracting a text feature from the search information and a content feature from each of the plurality pieces of media content comprises: obtaining a pre-trained neural network model, the pre-trained neural network model comprising a text encoder and a content encoder(Yuan para 0100 discloses using pre-trained encoder to obtain features “In some embodiments, a cross-modal encoder is pre-trained to encode the search image to obtain a search image representation. The cross-modal encoder encodes the text phrase to obtain a text phrase representation. The cross-modal encoder then selects the search image by comparing the search image representation and the text phrase representation”);
extracting the text feature from the search information by using the text encoder(Yuan, para 0080 discloses extracting text features from search or query information using text encoder “text phrase 600 is input to a text encoder to generate text features 610. As an example, text phrase 600 states “a couple of baseball players on a field”, which is a query from user 100”); and extracting the content feature from each of the plurality pieces of media content by using the content encoder(Yuan, para 0025 discloses encoding content features of candidate images “a cross-modal encoder of the image generation apparatus encodes candidate search images to obtain respective search image representation for each candidate search image”).
Regarding claim 3, Yuan, Liu and Sahu teach all the limitations of claim 1 and Liu further teaches wherein the transforming the plurality of content features to multiple mapped features comprises(Liu, para 0058 discloses transforming obtained feature through mapping “The computer device can map an obtained semantic feature to the instance feature space through the mapping relationship between the semantic feature space and the instance feature space, to obtain a transformed feature……”):
Sahu teaches performing feature mapping on the plurality of content features by using preset feature semantic distribution parameters, to obtain the mapped features(Sahu, para 0088 discloses mapping obtained features with target features based on threshold distance of matching “the query reformulation model 530 generates a feature vector for the provided autosuggest query and determines if the feature vector maps within a threshold distance to a feature vector of a known autosuggest query within the learned vector space”).
Regarding claim 6, Yuan, Liu and Sahu teach all the limitations of claim 1 and Yuan further teaches wherein the performing semantic recognition on the mapped features based on the text feature(Yuan, para 0103 discloses image/content features are compared for recognition based semantic information or feature in the query text “the target image includes substantially similar semantic features to semantic information as indicated in the text phrase”), to determine semantic types corresponding to the mapped features comprises: combining the text feature and a set formed by the mapped features obtained through mapping(Yuan, para 0082 discloses image and text features “output from text attention network 630 and output from image attention network 635 are input to image generation network 640 along with the upsampled noisy text features”),
to obtain a feature sequence; performing global attention processing on any mapped feature based on the feature sequence, to obtain a target feature corresponding to the mapped feature(Yuan, para 0078-0079 disclose performing attention processing “text attention network 630 is applied to text features 610 to obtain text attention features, where the target image 645 is generated based on the text attention features. Text attention network 630 is configured to generate text attention features…… image attention network 635 is applied to the search image to obtain image attention features”);
Liu teaches grouping/classifying features and classifying the target feature corresponding to the mapped feature, to obtain a semantic type corresponding to the mapped feature(Liu, para 0089 discloses grouping/clustering features by according to their closeness in semantic distance “the computer device may perform clustering with each point in the point cloud as a central point according to the feature distances between the semantic-fused instance features of the points in the point cloud, and find a plurality of adjacent points (including the central point itself) adjacent to each point. The plurality of adjacent points form a set of points, and a set of points may be considered as a local point cloud”).
Regarding claim 8, Yuan, Liu and Sahu teach all the limitations of claim 1 and Sahu further teaches wherein the determining search results for the search information from the media resource according to the target mapped features comprises: determining target media content corresponding to the target mapped features from the media resource; and adding the target media content to a search list, to obtain the search results for the search information (Sahu, element 310 of Fig. 3 and para 0048 disclose obtaining result based on target or newly mapped features (auto suggested query) where the results list a possible candidate results “In the first path corresponding to search results 310, the flow 300 includes act 312 of an autosuggest query being selected. For example, a user selects one of the autosuggest queries. In response, the autosuggest query system updates the first user interface to display the search results corresponding to the selection”; where Yuan teaches media contents).
Claim 10, Yuan teaches A computer device, comprising a processor and a memory, the memory having a plurality of instructions stored therein; and the processor, by executing the instructions from the memory, causing the computer device to perform a content search method including(Yuan, Fig. 4 discloses a computing device comprising memories, processor etc.): obtaining search information and a media resource, the media resource comprising a plurality of pieces of media content(Yuan, para 0004 discloses searching media/image based on extracted features “Embodiments of the inventive concept involve doing an image search based on the text prompt, and then using one or more retrieved images along with the text to generate a new image …….”); extracting a text feature from the search information (Yuan, para 0080 discloses extracting text features from search or query “text phrase 600 is input to a text encoder to generate text features 610. As an example, text phrase 600 states “a couple of baseball players on a field”, which is a query from user 100”) and a content feature from each of the plurality pieces of media content(Yuan, para 0079 discloses extraction content features from images “image attention network 635 is applied to the search image to obtain image attention features”); performing semantic recognition on the mapped features based on the text feature, to determine semantic types corresponding to the mapped features(Yuan, para 0103 discloses image/content features are compared for recognition based semantic information or feature in the query text “the target image includes substantially similar semantic features to semantic information as indicated in the text phrase”);
But Yuan does not explicitly teach transforming the plurality of content features to multiple mapped features, wherein a distance between a pair of mapped features represents semantic relevance between the pair of mapped features; grouping the mapped features corresponding to the same semantic type into a same combination; determining target mapped features meeting a relevance condition from different combinations based on the distances between mapped features in the different combinations; and determining search results for the search information from the media resource according to the target mapped features.
However, in the same field of endeavor of content feature analysis Liu teaches transforming the plurality of content features to multiple mapped features(Liu, para 0058 discloses transforming obtained feature through mapping “The computer device can map an obtained semantic feature to the instance feature space through the mapping relationship between the semantic feature space and the instance feature space, to obtain a transformed feature……”), wherein a distance between a pair of mapped features represents semantic relevance between the pair of mapped features(Liu, para 0079 discloses further discloses distances between features are Euclidean distance (similarity distance) “The feature distances are vector distances between the semantic-fused instance feature vector of each point and semantic-fused instance feature vectors of other points in the point cloud except each point. For example, the vector distances herein may be Euclidean distances or other distances”); grouping the mapped features corresponding to the same semantic type into a same combination(Liu, para 0089 discloses grouping/clustering features by according to their closeness in semantic distance “the computer device may perform clustering with each point in the point cloud as a central point according to the feature distances between the semantic-fused instance features of the points in the point cloud, and find a plurality of adjacent points (including the central point itself) adjacent to each point. The plurality of adjacent points form a set of points, and a set of points may be considered as a local point cloud”);
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate the feature of clustering content features based on their similarity of Liu into extraction of features of Liu to produce an expected result of mapping two sets of features based on their similarity. The modification would be obvious because one of ordinary skill in the art would be motivated to accurately categorize content features based on their semantic similarity distance (Liu, para 0089) .
But Yuan and Liu don’t explicitly teach determining target mapped features meeting a relevance condition from different combinations based on the distances between mapped features in the different combinations; and determining search results for the search information from the media resource according to the target mapped features;
However, in the same field of endeavor of content feature analysis Sahu teaches determining target mapped features meeting a relevance condition from different combinations based on the distances between mapped features in the different combinations(Sahu, para 0088 discloses mapping obtained features with target features based on threshold distance of matching “the query reformulation model 530 generates a feature vector for the provided autosuggest query and determines if the feature vector maps within a threshold distance to a feature vector of a known autosuggest query within the learned vector space”); and determining search results for the search information from the media resource according to the target mapped features(Sahu, element 310 of Fig. 3 and para 0048 disclose obtaining result based on target or newly mapped features (auto suggested query) “In the first path corresponding to search results 310, the flow 300 includes act 312 of an autosuggest query being selected. For example, a user selects one of the autosuggest queries. In response, the autosuggest query system updates the first user interface to display the search results corresponding to the selection”).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate the feature of mapping based on similarity distance of Sahu into mapping of content features of Yuan and Liu to produce an expected result of mapping two sets of features based on their similarity. The modification would be obvious because one of ordinary skill in the art would be motivated to improve search computational efficiency and accuracy (Sahu, abstract) .
Regarding claim 11, Yuan, Liu and Sahu teach all the limitations of claim 10 and Yuan further teaches wherein the extracting a text feature from the search information and a content feature from each of the plurality pieces of media content comprises: obtaining a pre-trained neural network model, the pre-trained neural network model comprising a text encoder and a content encoder (Yuan para 0100 discloses using pre-trained encoder to obtain features “In some embodiments, a cross-modal encoder is pre-trained to encode the search image to obtain a search image representation. The cross-modal encoder encodes the text phrase to obtain a text phrase representation. The cross-modal encoder then selects the search image by comparing the search image representation and the text phrase representation”);
extracting the text feature from the search information by using the text encoder (Yuan, para 0080 discloses extracting text features from search or query information using text encoder “text phrase 600 is input to a text encoder to generate text features 610. As an example, text phrase 600 states “a couple of baseball players on a field”, which is a query from user 100”); and extracting the content feature from each of the plurality pieces of media content by using the content encoder (Yuan, para 0025 discloses encoding content features of candidate images “a cross-modal encoder of the image generation apparatus encodes candidate search images to obtain respective search image representation for each candidate search image”).
Regarding claim 12, Yuan, Liu and Sahu teach all the limitations of claim 10 and Liu further teaches wherein the transforming the plurality of content features to multiple mapped features comprises(Liu, para 0058 discloses transforming obtained feature through mapping “The computer device can map an obtained semantic feature to the instance feature space through the mapping relationship between the semantic feature space and the instance feature space, to obtain a transformed feature……”):
Sahu teaches performing feature mapping on the plurality of content features by using preset feature semantic distribution parameters, to obtain the mapped features (Sahu, para 0088 discloses mapping obtained features with target features based on threshold distance of matching “the query reformulation model 530 generates a feature vector for the provided autosuggest query and determines if the feature vector maps within a threshold distance to a feature vector of a known autosuggest query within the learned vector space”).
Regarding claim 15, Yuan, Liu and Sahu teach all the limitations of claim 10, Yuan further teaches wherein the performing semantic recognition on the mapped features based on the text feature (Yuan, para 0103 discloses image/content features are compared for recognition based semantic information or feature in the query text “the target image includes substantially similar semantic features to semantic information as indicated in the text phrase”), to determine semantic types corresponding to the mapped features comprises: combining the text feature and a set formed by the mapped features obtained through mapping (Yuan, para 0082 discloses image and text features “output from text attention network 630 and output from image attention network 635 are input to image generation network 640 along with the upsampled noisy text features”),
to obtain a feature sequence; performing global attention processing on any mapped feature based on the feature sequence, to obtain a target feature corresponding to the mapped feature (Yuan, para 0078-0079 disclose performing attention processing “text attention network 630 is applied to text features 610 to obtain text attention features, where the target image 645 is generated based on the text attention features. Text attention network 630 is configured to generate text attention features…… image attention network 635 is applied to the search image to obtain image attention features”);
Liu teaches grouping/classifying features and classifying the target feature corresponding to the mapped feature, to obtain a semantic type corresponding to the mapped feature(Liu, para 0089 discloses grouping/clustering features by according to their closeness in semantic distance “the computer device may perform clustering with each point in the point cloud as a central point according to the feature distances between the semantic-fused instance features of the points in the point cloud, and find a plurality of adjacent points (including the central point itself) adjacent to each point. The plurality of adjacent points form a set of points, and a set of points may be considered as a local point cloud”).
Regarding claim 17, Yuan, Liu and Sahu teach all the limitations of claim 10 and Sahu further teaches wherein the determining search results for the search information from the media resource according to the target mapped features comprises: determining target media content corresponding to the target mapped features from the media resource; and adding the target media content to a search list, to obtain the search results for the search information (Sahu, element 310 of Fig. 3 and para 0048 disclose obtaining result based on target or newly mapped features (auto suggested query) where the results list a possible candidate results “In the first path corresponding to search results 310, the flow 300 includes act 312 of an autosuggest query being selected. For example, a user selects one of the autosuggest queries. In response, the autosuggest query system updates the first user interface to display the search results corresponding to the selection”; where Yuan teaches media contents),
Claim 19, Yuan teaches A non-transitory computer-readable storage medium, having a plurality of instructions stored therein, the instructions, when executed by a processor of a computer device, causing the computer device to perform a content search method including(Yuan, Fig. 4 & para 0066 disclose computer-readable storage media and processors” Computer-readable media includes both non-transitory computer storage media and communication media including any medium that facilitates transfer of code or data”): obtaining search information and a media resource, the media resource comprising a plurality of pieces of media content (Yuan, para 0004 discloses searching media/image based on extracted features “Embodiments of the inventive concept involve doing an image search based on the text prompt, and then using one or more retrieved images along with the text to generate a new image …….”); extracting a text feature from the search information (Yuan, para 0080 discloses extracting text features from search or query “text phrase 600 is input to a text encoder to generate text features 610. As an example, text phrase 600 states “a couple of baseball players on a field”, which is a query from user 100”) and a content feature from each of the plurality pieces of media content (Yuan, para 0079 discloses extraction content features from images “image attention network 635 is applied to the search image to obtain image attention features”);
performing semantic recognition on the mapped features based on the text feature, to determine semantic types corresponding to the mapped features (Yuan, para 0103 discloses image/content features are compared for recognition based semantic information or feature in the query text “the target image includes substantially similar semantic features to semantic information as indicated in the text phrase”);
But Yuan does not explicitly teach transforming the plurality of content features to multiple mapped features, wherein a distance between a pair of mapped features represents semantic relevance between the pair of mapped features; grouping the mapped features corresponding to the same semantic type into a same combination; determining target mapped features meeting a relevance condition from different combinations based on the distances between mapped features in the different combinations; and determining search results for the search information from the media resource according to the target mapped features.
However, in the same field of endeavor of content feature analysis Liu teaches transforming the plurality of content features to multiple mapped features(Liu, para 0058 discloses transforming obtained feature through mapping “The computer device can map an obtained semantic feature to the instance feature space through the mapping relationship between the semantic feature space and the instance feature space, to obtain a transformed feature……”), wherein a distance between a pair of mapped features represents semantic relevance between the pair of mapped features(Liu, para 0079 discloses further discloses distances between features are Euclidean distance (similarity distance) “The feature distances are vector distances between the semantic-fused instance feature vector of each point and semantic-fused instance feature vectors of other points in the point cloud except each point. For example, the vector distances herein may be Euclidean distances or other distances”); grouping the mapped features corresponding to the same semantic type into a same combination(Liu, para 0089 discloses grouping/clustering features by according to their closeness in semantic distance “the computer device may perform clustering with each point in the point cloud as a central point according to the feature distances between the semantic-fused instance features of the points in the point cloud, and find a plurality of adjacent points (including the central point itself) adjacent to each point. The plurality of adjacent points form a set of points, and a set of points may be considered as a local point cloud”);
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate the feature of clustering content features based on their similarity of Liu into extraction of features of Liu to produce an expected result of mapping two sets of features based on their similarity. The modification would be obvious because one of ordinary skill in the art would be motivated to accurately categorize content features based on their semantic similarity distance (Liu, para 0089) .
But Yuan and Liu don’t explicitly teach determining target mapped features meeting a relevance condition from different combinations based on the distances between mapped features in the different combinations; and determining search results for the search information from the media resource according to the target mapped features.
However, in the same field of endeavor of content feature analysis Sahu teaches determining target mapped features meeting a relevance condition from different combinations based on the distances between mapped features in the different combinations(Sahu, para 0088 discloses mapping obtained features with target features based on threshold distance of matching “the query reformulation model 530 generates a feature vector for the provided autosuggest query and determines if the feature vector maps within a threshold distance to a feature vector of a known autosuggest query within the learned vector space”); and determining search results for the search information from the media resource according to the target mapped features(Sahu, element 310 of Fig. 3 and para 0048 disclose obtaining result based on target or newly mapped features (auto suggested query) “In the first path corresponding to search results 310, the flow 300 includes act 312 of an autosuggest query being selected. For example, a user selects one of the autosuggest queries. In response, the autosuggest query system updates the first user interface to display the search results corresponding to the selection”).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate the feature of mapping based on similarity distance of Sahu into mapping of content features of Yuan and Liu to produce an expected result of mapping two sets of features based on their similarity. The modification would be obvious because one of ordinary skill in the art would be motivated to improve search computational efficiency and accuracy (Sahu, abstract) .
Regarding claim 20, Yuan, Liu and Sahu teach all the limitations of claim 19 and Yuan further teaches wherein the extracting a text feature from the search information and a content feature from each of the plurality pieces of media content comprises: obtaining a pre-trained neural network model, the pre-trained neural network model comprising a text encoder and a content encoder (Yuan para 0100 discloses using pre-trained encoder to obtain features “In some embodiments, a cross-modal encoder is pre-trained to encode the search image to obtain a search image representation. The cross-modal encoder encodes the text phrase to obtain a text phrase representation. The cross-modal encoder then selects the search image by comparing the search image representation and the text phrase representation”);
extracting the text feature from the search information by using the text encoder (Yuan, para 0080 discloses extracting text features from search or query information using text encoder “text phrase 600 is input to a text encoder to generate text features 610. As an example, text phrase 600 states “a couple of baseball players on a field”, which is a query from user 100”); and extracting the content feature from each of the plurality pieces of media content by using the content encoder (Yuan, para 0025 discloses encoding content features of candidate images “a cross-modal encoder of the image generation apparatus encodes candidate search images to obtain respective search image representation for each candidate search image”).
Claim 4 and 13 are rejected under 35 U.S.C. 103 as being unpatentable over Yuan, Xin et al (PGPUB Document No. 20230260164), hereafter referred as to “Yuan”, in view of Liu, Shu et al (PGPUB Document No. 20210350183), hereafter, referred to as “Liu”, in view of Sahu, Tezan et al (PGPUB Document No. 20240411824), hereafter, referred to as “Sahu”, in view of Tse, Kwong et al (US Patent No. 11592371), hereafter, referred to as “Tse”.
Regarding claim 4, Yuan, Liu and Sahu teach all the limitations of claim 3 but don’t explicitly teach wherein the feature semantic distribution parameters comprise non-linear distribution parameters and linear distribution parameters, and the performing feature mapping on the plurality of content features by using a preset feature semantic distribution parameters, to obtain the mapped features comprises:
performing non-linear transformation on the plurality of content features by using the preset non-linear distribution parameters, to obtain intermediate features; and performing linear transformation on the intermediate features by using the preset linear distribution parameters, to obtain the mapped features.
However, in the same field of endeavor of content feature analysis Sahu teaches wherein the feature semantic distribution parameters comprise non-linear distribution parameters and linear distribution parameters, and the performing feature mapping on the plurality of content features by using a preset feature semantic distribution parameters, to obtain the mapped features comprises:
performing non-linear transformation on the plurality of content features by using the preset non-linear distribution parameters, to obtain intermediate features; and performing linear transformation on the intermediate features by using the preset linear distribution parameters, to obtain the mapped features(Tse, in col 27:65~col 28:1-3 discloses determining correlation between parameters by linear and non-linear distribution, this disclosed teaching of correlating parameters can be applied for mapping/correlating features based on semantic parameter distribution “determining a correlation between parameters, determining a variance and/or standard deviation of the set (or a subset) of the event parameters, using an equation relating the event parameters to the population parameters, using a weighted average of the event parameters, determining a regression (e.g., linear regression, nonlinear regression, etc.) between the distribution of event and/or frame parameters, and/or can be otherwise determined” ).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate the feature of considering parameter distribution for mapping data/features of Tse into mapping of content features of Yuan, Liu and Sahu to produce an expected result of mapping two sets of features based on their similarity. The modification would be obvious because one of ordinary skill in the art would be motivated to use linear/non-linear feature data distribution for reliable mapping of features (Tse, col 27:65~col 28:1-3).
Regarding claim 13, Yuan, Liu and Sahu teach all the limitations of claim 12 but don’t explicitly teach wherein the feature semantic distribution parameters comprise non-linear distribution parameters and linear distribution parameters, and the performing feature mapping on the plurality of content features by using a preset feature semantic distribution parameters, to obtain the mapped features comprises: performing non-linear transformation on the plurality of content features by using the preset non-linear distribution parameters, to obtain intermediate features; and performing linear transformation on the intermediate features by using the preset linear distribution parameters, to obtain the mapped features.
However, in the same field of endeavor of content feature analysis Sahu teaches wherein the feature semantic distribution parameters comprise non-linear distribution parameters and linear distribution parameters, and the performing feature mapping on the plurality of content features by using a preset feature semantic distribution parameters, to obtain the mapped features comprises:
performing non-linear transformation on the plurality of content features by using the preset non-linear distribution parameters, to obtain intermediate features; and performing linear transformation on the intermediate features by using the preset linear distribution parameters, to obtain the mapped features(Tse, in col 27:65~col 28:1-3 discloses determining correlation between parameters by linear and non-linear distribution, this disclosed teaching of correlating parameters can be applied for mapping/correlating features based on semantic parameter distribution “determining a correlation between parameters, determining a variance and/or standard deviation of the set (or a subset) of the event parameters, using an equation relating the event parameters to the population parameters, using a weighted average of the event parameters, determining a regression (e.g., linear regression, nonlinear regression, etc.) between the distribution of event and/or frame parameters, and/or can be otherwise determined” ).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate the feature of considering parameter distribution for mapping data/features of Tse into mapping of content features of Yuan, Liu and Sahu to produce an expected result of mapping two sets of features based on their similarity. The modification would be obvious because one of ordinary skill in the art would be motivated to use linear/non-linear feature data distribution for reliable mapping of features (Tse, col 27:65~col 28:1-3) .
Claim 5 and 14 are rejected under 35 U.S.C. 103 as being unpatentable over Yuan, Xin et al (PGPUB Document No. 20230260164), hereafter referred as to “Yuan”, in view of Liu, Shu et al (PGPUB Document No. 20210350183), hereafter, referred to as “Liu”, in view of Sahu, Tezan et al (PGPUB Document No. 20240411824), hereafter, referred to as “Sahu”, in view of JI, Chao et al (PGPUB Document No. 20250054280), hereafter, referred to as “JI”.
Regarding claim 5, Yuan, Liu and Sahu teach all the limitations of claim 3 but don’t explicitly teach further comprising: obtaining a training sample set and initial distribution parameters, the training sample set comprising positive samples and negative samples; and updating the initial distribution parameters through contrastive learning with the positive samples and the negative samples, to obtain the feature semantic distribution parameters.
However, in the same field of endeavor of AI model training JI teaches further comprising: obtaining a training sample set and initial distribution parameters, the training sample set comprising positive samples and negative samples; and updating the initial distribution parameters through contrastive learning with the positive samples and the negative samples, to obtain the feature semantic distribution parameters(Ji, para 0082 discloses model training with positive and negative sample for contrastive learning and following disclosed teachings can similarly be applied to distribution parameters “the image-text matching model can be trained by using the positive and negative samples and in a manner of contrastive learning, so it is convenient to increase a number of negative samples, so as to increase a number of samples to improve training effect of the image-text matching model”).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate the feature of training a learning model with positive and negative samples of JI into mapping of content features of Yuan, Liu and Sahu to produce an expected result of mapping two sets of features based on their similarity. The modification would be obvious because one of ordinary skill in the art would be motivated to improve the training effect in matching using contrastive learning (Ji, para 0065) .
Regarding claim 14, Yuan, Liu and Sahu teach all the limitations of claim 10 but don’t explicitly teach wherein the method further comprises: obtaining a training sample set and initial distribution parameters, the training sample set comprising positive samples and negative samples; and updating the initial distribution parameters through contrastive learning with the positive samples and the negative samples, to obtain the feature semantic distribution parameters.
However, in the same field of endeavor of AI model training JI teaches wherein the method further comprises: obtaining a training sample set and initial distribution parameters, the training sample set comprising positive samples and negative samples; and updating the initial distribution parameters through contrastive learning with the positive samples and the negative samples, to obtain the feature semantic distribution parameters (Ji, para 0082 discloses model training with positive and negative sample for contrastive learning and following disclosed teachings can similarly be applied to distribution parameters “the image-text matching model can be trained by using the positive and negative samples and in a manner of contrastive learning, so it is convenient to increase a number of negative samples, so as to increase a number of samples to improve training effect of the image-text matching model”).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate the feature of training a learning model with positive and negative samples of JI into mapping of content features of Yuan, Liu and Sahu to produce an expected result of mapping two sets of features based on their similarity. The modification would be obvious because one of ordinary skill in the art would be motivated to improve the training effect in matching using contrastive learning (Ji, para 0065).
Claim 7 and 16 are rejected under 35 U.S.C. 103 as being unpatentable over Yuan, Xin et al (PGPUB Document No. 20230260164), hereafter referred as to “Yuan”, in view of Liu, Shu et al (PGPUB Document No. 20210350183), hereafter, referred to as “Liu”, in view of Sahu, Tezan et al (PGPUB Document No. 20240411824), hereafter, referred to as “Sahu”, in view of Wang, Nan et al (PGPUB Document No. 20240273010), hereafter, referred to as “Wang”.
Regarding claim 7, Yuan, Liu and Sahu teach all the limitations of claim 1 and Sahu further teaches wherein the determining target mapped features meeting a relevance condition from different combinations based on the distances between mapped features in the different combinations comprises(Sahu, para 0088 discloses mapping obtained features with target features based on threshold distance of matching “the query reformulation model 530 generates a feature vector for the provided autosuggest query and determines if the feature vector maps within a threshold distance to a feature vector of a known autosuggest query within the learned vector space”):
Sahu further teaches determining, in each combination, a ranking number of similarity between each mapped feature and the text feature(Sahu, para 0075 discloses determining matching number of similarity in feature “In some implementations, the query gateway system 206 determines a match based on a threshold number or amount of matching words. In various implementations, the AI chat eligibility model 520 is a machine-learning model that determines a match based on the proximity of the autosuggest query to an eligible autosuggest query within the vector space”);
But Yuan, Liu and Sahu don’t explicitly teach and selecting, from each combination, the mapped features with ranking numbers not exceeding a preset number as the target mapped features.
However, in the same field of endeavor of feature mapping Wang teaches and selecting, from each combination, the mapped features with ranking numbers not exceeding a preset number as the target mapped features(Wang, para 0010 discloses number of mapped features on both sides are equal “The number of reference program errors may be a maximum value selected from a set of the numbers of program errors mapped to each test case across features. The number of reference features may be a maximum value selected from a set of the numbers of features mapped to each test case across features”; here the examiner interprets inconsistent with specification and as it would been interpreted by one ordinary skilled in the art to mean number of mapped features on both sides are equal).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate the feature of mapping a certain number of features of Wang into mapping of content features of Yuan, Liu and Sahu to produce an expected result of mapping two sets of features based on their similarity. The modification would be obvious because one of ordinary skill in the art would be motivated to map source to target features not more than number of features in the target to avoid generating unnecessary source features(Wang, para 0010) .
Regarding claim 16, Yuan, Liu and Sahu teach all the limitations of claim 10 and Sahu further teaches wherein the determining target mapped features meeting a relevance condition from different combinations based on the distances between mapped features in the different combinations comprises (Sahu, para 0088 discloses mapping obtained features with target features based on threshold distance of matching “the query reformulation model 530 generates a feature vector for the provided autosuggest query and determines if the feature vector maps within a threshold distance to a feature vector of a known autosuggest query within the learned vector space”):
Sahu further teaches determining, in each combination, a ranking number of similarity between each mapped feature and the text feature (Sahu, para 0075 discloses determining matching number of similarity in feature “In some implementations, the query gateway system 206 determines a match based on a threshold number or amount of matching words. In various implementations, the AI chat eligibility model 520 is a machine-learning model that determines a match based on the proximity of the autosuggest query to an eligible autosuggest query within the vector space”);
But Yuan, Liu and Sahu don’t explicitly teach and selecting, from each combination, the mapped features with ranking numbers not exceeding a preset number as the target mapped features.
However, in the same field of endeavor of feature mapping Wang teaches and selecting, from each combination, the mapped features with ranking numbers not exceeding a preset number as the target mapped features (Wang, para 0010 discloses number of mapped features on both sides are equal “The number of reference program errors may be a maximum value selected from a set of the numbers of program errors mapped to each test case across features. The number of reference features may be a maximum value selected from a set of the numbers of features mapped to each test case across features”; here the examiner interprets inconsistent with specification and as it would been interpreted by one ordinary skilled in the art to mean number of mapped features on both sides are equal).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate the feature of mapping a certain number of features of Wang into mapping of content features of Yuan, Liu and Sahu to produce an expected result of mapping two sets of features based on their similarity. The modification would be obvious because one of ordinary skill in the art would be motivated to map source to target features not more than number of features in the target to avoid generating unnecessary source features(Wang, para 0010) .
Claim 9 and 18 are rejected under 35 U.S.C. 103 as being unpatentable over Yuan, Xin et al (PGPUB Document No. 20230260164), hereafter referred as to “Yuan”, in view of Liu, Shu et al (PGPUB Document No. 20210350183), hereafter, referred to as “Liu”, in view of Sahu, Tezan et al (PGPUB Document No. 20240411824), hereafter, referred to as “Sahu”, in view of Zhang, Renhui et al (PGPUB Document No. 20190220475), hereafter, referred to as “Zhang”.
Regarding claim 9, Yuan, Liu and Sahu teach all the limitations of claim 8 and Sahu further teaches wherein the adding the target media content to a search list, to obtain the search results for the search information comprises: adding the target media content to the search list, to obtain an updated search list(Sahu, element 310 of Fig. 3 and para 0048 disclose obtaining result based on target or newly mapped features (auto suggested query) where the results list a possible candidate results “In the first path corresponding to search results 310, the flow 300 includes act 312 of an autosuggest query being selected. For example, a user selects one of the autosuggest queries. In response, the autosuggest query system updates the first user interface to display the search results corresponding to the selection”; where Yuan teaches media contents);
But Yuan, Liu and Sahu don’t explicitly teach and ranking the target media content in the updated search list according to search parameters, to obtain the search results for the search information, the search parameters comprising at least one of similarity between the target media content and the search information, visual quality, timestamp, and popularity.
However, in the same field of endeavor of feature mapping Wang teaches and ranking the target media content in the updated search list according to search parameters, to obtain the search results for the search information, the search parameters comprising at least one of similarity between the target media content and the search information, visual quality, timestamp, and popularity(Zhang, para 0088-0089 discloses ranking contents based on search information (topic) quality timestamp and popularity “Content items may be ranked (206) in any desired manner, such as based on one or more of: a popularity level of the respective content item, an age of the respective content item, a size of the respective content item, an author associated with the respective content item, a publisher associated with the respective content item, and a quality rating of the respective content item…. The age of the respective content item may be measured in any desired manner, such as based on a timestamp in an article containing the content item”).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate the feature of ranking contents of Zhang into mapping of content features of Yuan, Liu and Sahu to produce an expected result of getting a most related content items. The modification would be obvious because one of ordinary skill in the art would be motivated to identify best contents by ranking them(Zhang, para 0088).
Regarding claim 18, Yuan, Liu and Sahu teach all the limitations of claim 10 and Sahu further teaches wherein the adding the target media content to a search list, to obtain the search results for the search information comprises: adding the target media content to the search list, to obtain an updated search list (Sahu, element 310 of Fig. 3 and para 0048 disclose obtaining result based on target or newly mapped features (auto suggested query) where the results list a possible candidate results “In the first path corresponding to search results 310, the flow 300 includes act 312 of an autosuggest query being selected. For example, a user selects one of the autosuggest queries. In response, the autosuggest query system updates the first user interface to display the search results corresponding to the selection”; where Yuan teaches media contents);
But Yuan, Liu and Sahu don’t explicitly teach and ranking the target media content in the updated search list according to search parameters, to obtain the search results for the search information, the search parameters comprising at least one of similarity between the target media content and the search information, visual quality, timestamp, and popularity.
However, in the same field of endeavor of feature mapping Wang teaches and ranking the target media content in the updated search list according to search parameters, to obtain the search results for the search information, the search parameters comprising at least one of similarity between the target media content and the search information, visual quality, timestamp, and popularity (Zhang, para 0088-0089 discloses ranking contents based on search information (topic) quality timestamp and popularity “Content items may be ranked (206) in any desired manner, such as based on one or more of: a popularity level of the respective content item, an age of the respective content item, a size of the respective content item, an author associated with the respective content item, a publisher associated with the respective content item, and a quality rating of the respective content item…. The age of the respective content item may be measured in any desired manner, such as based on a timestamp in an article containing the content item”).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate the feature of ranking contents of Zhang into mapping of content features of Yuan, Liu and Sahu to produce an expected result of getting a most related content items. The modification would be obvious because one of ordinary skill in the art would be motivated to identify best contents by ranking them(Zhang, para 0088) .
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to ABDULLAH A DAUD whose telephone number is (469)295-9283. The examiner can normally be reached M~F: 9:30 am~6:30 pm.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Amy Ng can be reached at 571-270-1698. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of published or unpublished applications may be obtained from Patent Center. Unpublished application information in Patent Center is available to registered users. To file and manage patent submissions in Patent Center, visit: https://patentcenter.uspto.gov. Visit https://www.uspto.gov/patents/apply/patent-center for more information about Patent Center and https://www.uspto.gov/patents/docx for information about filing in DOCX format. For additional questions, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/ABDULLAH A DAUD/Examiner, Art Unit 2164
/AMY NG/Supervisory Patent Examiner, Art Unit 2164
Prosecution Insights