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
The information disclosure statement (IDS) submitted on 12/01/2025 is in compliance with the provisions of 37 CFR 1.97. Accordingly, the information disclosure statement is being considered by the examiner.
Response to Amendment
The amendments filed 11/21/2025 have been entered.
Claims 1-3, 6, 8-10, 12-14, 17, and 19 remain pending within the application.
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.
Claims 1-3, 6, 8-10, 12-14, 17, and 19 are rejected under 35 U.S.C. 103 as being unpatentable over Hill et al. (Pub. No.: US 2011/0320454 A1), hereafter Hill, in view of Srivastava et al. ("Bag of Tricks for Retail Product Image Classification"), hereafter Srivastava, in further view of Krestel et al. ("Latent Dirichlet Allocation for Tag Recommendation"), hereafter Krestel.
Regarding claim 1, Hill discloses:
A computer implemented method to classify an input, the method comprising (Hill, Fig. 7, element 701 and Fig. 2, step 204 classifies digital objects as input using the processor recited in Fig. 7),
receiving, by a classification platform processor from a user device via a user interface associated with a classification platform, an input comprising at least one image from a real world domain (Hill, paragraph 0037, lines 4-5 “Media items 114 are input to a classifier 116”, paragraph 0032, lines 5-9 "content is generated or otherwise obtained. Content may include any artifact including a photograph, digital image, video, graphic, etc. The content in one illustrative example includes a photo of dolphins in the ocean."),
applying, by the classification platform processor, a base model to the input … (Hill, Fig. 2, step 204 classifies digital objects as input by applying a base model),
predicting, by the classification platform processor, at least a first base concept associated with the input (Hill, Fig. 7, element 701 and Fig. 1A step 104 teaches predicting at least a first base concept associated with input 102),
querying, by the classification platform processor utilizing the at least first base concept, a mapping data structure for at least one … concept (Hill, Fig 2 and paragraph 0040, lines 8-11 “The categorization/classification from block 204 is refined in block 212 based on the faceted ontology properties or constraints from block 210. “ teaches refining based on faceted ontology properties as querying the mapping data structure in 208 for concepts using the base concepts learned in 204),
receiving, by the classification platform processor, a plurality of … concepts (Hill, paragraph [0041] and Fig 2 teaches receiving a plurality of concepts in 214),
mapping, by the classification platform processor utilizing mapping data generated during the training process and stored in a mapping database, the plurality of…concepts to the at least first base concept (Hill, Fig. 2, element 208, paragraph 0040, and paragraph 0082 teaches mapping multiple concepts to the base concept using a faceted taxonomy, which is mapping data generated during training process and stored in a database),
identifying, by the classification platform processor, an ignore list of concepts associated with the at least first base concept (Hill, Fig 1B and paragraph 0038, lines 1 "Category label refinement is provided in block 120" and line 5 "label selection for exclusion (filtering)," teaches an ignore list associated with the first base concept for filtering).
filtering, by the classification platform processor, the mapped plurality of… concepts to remove any … concepts on the ignore list and to keep the remaining … concepts having a confidence score greater than a threshold amount (Hill, Fig 1B and paragraph 0038, lines 1 "Category label refinement is provided in block 120", line 5 "label selection for exclusion (filtering),", Fig. 1A and paragraphs [0032-0033] and [0039] teaches filtering to remove concepts on the ignore list in step 120 of Fig. 1B and refining the mapped concepts as filtering the mapped concepts using the confidence scores of nodes, using a minimum or maximum range of classification confidence as the threshold),
outputting, by the classification platform processor to a database, the … concepts which were kept as a classification of the input (Hill, Fig. 1B teaches outputting the determined concept as classification in step 122).
Hill discloses applying a base model to the input, but does not disclose training the base model on images from a first domain different from the real world domain.
Srivastava discloses:
wherein … model was trained during a training process on images from a first domain different from the real world domain (Srivastava, page 4, Figs 6-7 and left col paragraph 1, lines 3-5 “We take in-vitro images of retail products as training data and insitu images were taken as testing data” teaches a model trained on images from a first domain different from the real world domain).
Hill and Srivastava are analogous art because they are from the same field of endeavor, machine learning and concept learning.
It would have been obvious for one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the teachings of Hill to include wherein … model was trained during a training process on images from a first domain different from the real world domain, based on the teachings of Srivastava. The motivation for doing so would have been “for classification to achieve better accuracy” (Srivastava, page 2, left col paragraph 1, line 4).
While Hill, in view of Srivastava, discloses querying, utilizing the at least first base concept, a mapping data structure for at least one … concept and receiving a plurality of … concepts, they not explicitly disclose these concepts to be custom concepts.
Krestel discloses:
querying…a mapping …for at least one custom concept (Krestel, Table 1 and page 62, section 2.1 Problem Definition, lines 5-8 “The goal of collective tag recommendation is to suggest new tags for a resource ri with only a few bookmarks based on tag assignments to other resources collected in Y = ….” Teaches querying a mapping of tag assignments for at least one custom concept, i.e. suggested tag).
receiving…custom concept (Krestel, page 62, section 2.1 Problem Definition, lines 6-7 “suggest new tags for a resource ri" teaches receiving custom concepts).
It would have been obvious for one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the teachings of Hill, in view of Srivastava, to include querying…a mapping …for at least one custom concept and custom concepts, based on the teachings of Krestel. The motivation for doing so would have been to achieve "better accuracy, ... recommends more specific tags, which are more useful for search." (Krestel, page 67, section 5 Conclusion and Future Work, paragraph 1, lines 3-5).
Regarding claim 2, Hill, in view of Srivastava, in further view of Krestel, discloses the computer-implemented method of claim 1…predicting at least a first base concept associated with the input. Hill further discloses:
generating a confidence score associated with the at least first base concept (Hill, Fig 1A step 104 and paragraph 0032, lines 14-15 "categories with confidence score," teaches generating confidence scores associated with the at least first base concept as demonstrated in Fig 1A step 104, such as generating “0.71” for at least "Dolphin")
Regarding claim 3, Hill, in view of Srivastava, in further view of Krestel, discloses the computer-implemented method of claim 1 … predicting at least a first base concept associated with the input. Hill further discloses:
generating a first confidence score associated with the at least first base concept (Hill, Fig 1A step 104 teaches generating a first confidence score associated with the at least first base concept, such as generating “0.71” for at least "Dolphin")
predicting at least a second base concept associated with the input (Hill, Fig 1A step 104 teaches predicting at least a second base concept associated with the input, such as "Blue")
generating a second confidence score associated with the at least second base concept (Hill, Fig 1A step 104 teaches generating a second confidence score associated with the at least second base concept, such as generating “0.87” for at least "Blue").
Regarding claim 6, Hill, in view of Srivastava, in view of Krestel, discloses the computer-implemented method of claim 5. Hill, in view of Krestel, discloses outputting custom concepts in claim 5. Hill further discloses:
ranking the… concepts which were kept to output the highest ranked concept first (Hill, paragraph 0059, lines 1-3 "Each of the re-ranking factors f may serve as re-scoring factor for the classifier confidence among the different tags within the same image." Teaches ranking concepts using re-ranking factor f to classify input image for output of highest ranked concept).
Hill, in view of Srivastava, discloses ranking the… concepts which were kept to output the highest ranked concept first, but does not explicitly disclose those concepts to be custom.
Krestel discloses:
custom concepts (Krestel, page 61, Section 1 Introduction, paragraph 4, lines 6-7 "Based on this, other tags belonging to the recommended topics can be recommended" teaches recommending other tags as custom concepts that are determined by mapping to recommended topics).
It would have been obvious for one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the teachings of Hill, in view of Srivastava, to include custom concepts, based on the teachings of Krestel. The motivation for doing so would have been to achieve "better accuracy, ... recommends more specific tags, which are more useful for search." (Krestel, page 67, section 5 Conclusion and Future Work, paragraph 1, lines 3-5).
Regarding claim 8, Hill, in view of Srivastava, in further view of Krestel, discloses the computer-implemented method of claim 1. Hill further discloses:
the mapping data structure includes a plurality of base concepts including the at least first base concept (Hill, paragraph 0039, lines 1-4 "Category label refinement may be based on properties of nodes (leaf or internal, domain-specific, number of siblings, number of descendants, depth in tree, etc.); confidence scores of nodes," and paragraph 0073, lines 4-6 "Images or artifacts are categorized. Each category represents a unique semantic idea and is represented as a node, called a category node." teaches using the image categories as the base concepts in the mapping data structure, which includes the categories for refinement, i.e., the first base concept)
Regarding claim 9, Hill, in view of Srivastava, in further view of Krestel, discloses the computer-implemented method of claim 8. Hill further discloses:
the mapping data structure includes, for each of the plurality of base concepts, information identifying one or more corresponding … concepts (Hill, paragraph 0073, lines 6-8 "Each category node is optionally linked to one or more children nodes that reflect semantic decomposition of the parent category." teaches children nodes as information identifying one or more corresponding concepts).
While Hill, in view of Srivastava, discloses the mapping data structure includes, for each of the plurality of base concepts, information identifying one or more corresponding … concepts, Hill does not explicitly disclose these concepts to be custom concepts.
Krestel discloses:
custom concepts (Krestel, page 61, Section 1 Introduction, paragraph 4, lines 6-7 "Based on this, other tags belonging to the recommended topics can be recommended" teaches recommending other tags as custom concepts that are determined by mapping to recommended topics).
It would have been obvious for one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the teachings of Hill, in view of Srivastava, to include custom concepts, based on the teachings of Krestel. The motivation for doing so would have been to achieve "better accuracy, ... recommends more specific tags, which are more useful for search." (Krestel, page 67, section 5 Conclusion and Future Work, paragraph 1, lines 3-5).
Regarding claim 10, Hill, in view of Srivastava, in further view of Krestel, discloses the computer-implemented method of claim 9. Hill further discloses:
the mapping data structure further includes, for the one or more corresponding .. concepts, a confidence score indicating a confidence in the relationship between the one or more corresponding … concepts and the associated base concept (Hill, Fig 2 and paragraph 0052, lines 6-8 "We can augment the set of tags by propagating tag confidence scores bottom-up in the taxonomy" teaches confidence scores indicating a confidence in the relationship between base and custom concepts in the taxonomy mapping structure of 208 in Fig. 2).
While Hill, in view of Srivastava, discloses the mapping data structure further includes, for the one or more corresponding .. concepts, a confidence score indicating a confidence in the relationship between the one or more corresponding … concepts and the associated base concept, they do not explicitly disclose these concepts to be custom concepts.
Krestel discloses:
custom concepts (Krestel, page 61, Section 1 Introduction, paragraph 4, lines 6-7 "Based on this, other tags belonging to the recommended topics can be recommended" teaches recommending other tags as custom concepts that are determined by mapping to recommended topics).
It would have been obvious for one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the teachings of Hill, in view of Srivastava, to include custom concepts, based on the teachings of Krestel. The motivation for doing so would have been to achieve "better accuracy, ... recommends more specific tags, which are more useful for search." (Krestel, page 67, section 5 Conclusion and Future Work, paragraph 1, lines 3-5).
Claim 12 is substantially similar to claim 1, but for the recitation of a mapping module operably connected to the classification platform processing unit (Hill, Fig. 7, element 708 and ¶[0083] teaches the refinement module as a mapping module operably connected to the classification platform processing unit), and thus claim 12 is rejected on the same basis as claim 1.
Regarding Claim 13, Hill, in view of Srivastava, in further view of Krestel, discloses the classification platform of claim 12. Hill further discloses:
the input is one of an image and a video (Hill, paragraph 0032, lines 6-8 "Content may include any artifact including a photograph, digital image, video, graphic, etc.").
Claims 14 is substantially similar to claim 3, and thus is rejected on the same basis as claim 3.
Claims 17 is substantially similar to claim 6, and thus is rejected on the same basis as claim 6.
Regarding claim 19, Hill, in view of Srivastava, in further view of Krestel, discloses the classification platform of claim 12. Srivastava further discloses:
the input is from a different domain than a set of inputs used to train … model (Srivastava, page 4, Figs 6-7 and left col paragraph 1, lines 3-5 “We take in-vitro images of retail products as training data and insitu images were taken as testing data” teaches a test insitu image as input, which is from a different domain than a set of in vitro inputs used to train the model).
It would have been obvious for one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the teachings of Hill to include the input is from a different domain than a set of inputs used to train … model, based on the teachings of Srivastava. The motivation for doing so would have been “for classification to achieve better accuracy” (Srivastava, page 2, left col paragraph 1, line 4).
Response to Arguments
Applicant's arguments filed 11/21/2025 have been fully considered with regards to the 35 U.S.C. 101 rejections, and they are found persuasive. The rejections are withdrawn.
Applicant's arguments filed 11/21/2025 have been fully considered with regards to the 35 U.S.C. 102/103 rejection, but they are not persuasive.
The applicant asserts on page 22 of the remarks “Thus, Applicant respectfully submits neither Hill nor Srivastava teaches or suggests a classification platform processor querying ... utilizing the at least first base concept, a mapping data structure for at least one custom concept then receiving ... a plurality of custom concepts before than mapping ... utilizing mapping data generated during the training process and stored in a mapping database, the plurality of custom concepts to the at least first base concept as now required by claims 1 and 12.” The Examiner respectfully disagrees, as the amended limitations to claim 1 in question mostly incorporate the limitations of the previously rejected claim 7. Hill teaches querying, by utilizing the at least first base concept, a mapping data structure for at least one concept by refining based on faceted ontology properties as querying the mapping data structure in 208 for concepts using the base concepts learned in 204 (Hill, Fig 2 and paragraph 0040, lines 8-11), as well as receiving a plurality of concepts by receiving a plurality of concepts in Fig. 2 element 214 (Hill, paragraph [0041] and Fig 2). Krestel discloses querying a mapping for at least one custom concept by querying a mapping of tag assignments for at least one custom concept, i.e. suggested tag (Krestel, Table 1 and page 62, section 2.1 Problem Definition, lines 5-8 “The goal of collective tag recommendation is to suggest new tags for a resource ri with only a few bookmarks based on tag assignments to other resources collected in Y = ….”), and receiving custom concept (Krestel, page 62, section 2.1 Problem Definition, lines 6-7). It would have been obvious for one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the teachings of Hill, in view of Srivastava, to include querying a mapping for at least one custom concept and custom concepts, based on the teachings of Krestel. The motivation for doing so would have been to achieve "better accuracy, ... recommends more specific tags, which are more useful for search." (Krestel, page 67, section 5 Conclusion and Future Work, paragraph 1, lines 3-5).
The applicant asserts on page 23 of the remarks “Applicant notes that the cited sentence of Krestel which allegedly discloses "custom concepts," namely the sentence on page 61, Section 1, paragraph 4 under the "Introduction" section, instead recites the following: "Based on this, other tags belonging to the recommended topics can be recommended." The context in which this sentence appears is one in which overcoming the cold start problem for tagging new resources is the goal, and to utilize LDA to do so. Applicant respectfully submits that such disclosure is not equivalent to having a classification platform processor .. utilizing the at least first base concept, a mapping data structure for at least one custom concept then receive ... a plurality of custom concepts before than mapping ... utilizing mapping data generated during the training process and stored in a mapping database, the plurality of custom concepts to the at least first base concept, as now required by claims 1 and 12.”. The Examiner respectfully disagrees, as Krestel does not need to disclose features already taught by Hill, in view of Srivastava. The idea that the custom concepts taught by Krestel overcome the cold start problem for tagging new resources is not relevant. In response to applicant's arguments against the Krestel reference individually, one cannot show nonobviousness by attacking references individually where the rejections are based on combinations of references. See In re Keller, 642 F.2d 413, 208 USPQ 871 (CCPA 1981); In re Merck & Co., 800 F.2d 1091, 231 USPQ 375 (Fed. Cir. 1986).
Conclusion
Applicant's amendment necessitated the new ground(s) of rejection presented in this Office action. Accordingly, THIS ACTION IS MADE FINAL. See MPEP § 706.07(a). Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any extension fee pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to HUMAIRA ZAHIN MAUNI whose telephone number is (703)756-5654. The examiner can normally be reached Monday - Friday, 9 am - 5 pm (ET).
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, MATT ELL can be reached at (571) 270-3264. 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.
/H.Z.M./Examiner, Art Unit 2141
/MATTHEW ELL/Supervisory Patent Examiner, Art Unit 2141