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 .
Acknowledgement
Acknowledgement is made of applicant’s amendment made on 08/21/2026. Applicant’s submission filed has been entered and made of record.
Status of the Claims
Claims 1-20 are pending.
Response to Applicant’s Arguments
In response to “Nowhere does Zhao disclose determining any score based on a number of times an entity is referenced in content, and nowhere does Zhao disclose counting references to an entity for any purpose”.
In view of such amendment to the independent claims, anticipation rejection under Zhao has been withdrawn. Upon further search and consideration, please see details of a new combination of references set forth below.
In response to “The candidate entities to which the embedding scores of Zhao are assigned are entities that are missing from the triple under consideration and are predicted by a prediction model, rather than entities identified as being referenced in received content. (See id. at paragraph 41 on page 4 and paragraph 50 on page 5.)” and “The first entity of Zhao is a known entity that is already represented in the knowledge graph, and the only entity that Zhao obtains is the second entity missing from an incomplete triple. (See id. at paragraph 41 on page 4 and paragraph 50 on page 5.) Nowhere does Zhao disclose mapping an entity identified as being referenced in received content to an entry in a knowledge base, and nowhere does Zhao disclose providing an entity mapping as a result of any such operation. (See generally Zhao.) Consequently, Applicant respectfully submits that Zhao also fails to disclose the limitation: "map, using the first ML model trained as the mapping agent, each of the one or more entities to respective one or more entries in a knowledge base to provide one or more entity mappings," affirmatively required by currently amended independent claim 1, and similarly required by currently amended independent claim 11”.
According to Zhao, a knowledge graph is a graph based data structure in which nodes represent entities and an edge between two nodes represents relationship between entities (p. 1, ¶10; i.e., an entry in the knowledge graph comprises two entities and an edge relationship or triple) where text modality is used to represent entity text information in the knowledge graph and image modality is used to represent entity image information in the knowledge graph (p. 5, ¶52).
Zhao at Fig 5 shows an exemplary application of knowledge graph completion of receiving content comprising: (1) Text modality introducing the entity as a certain athlete, (2) the structural modality containing athlete’s unimodal relationships including nationality (athlete having nationalities “Country A” and “Country B”, while athlete’s friends C and D having “Country B” nationality), and (3) the image modality showing the scene of athlete playing (p. 13, ¶133; i.e., receiving content including image and text):
PNG
media_image1.png
364
685
media_image1.png
Greyscale
In this example, the text modality introduced the entity as a certain athlete and the image modality shows the scene of the athlete playing (p. 13, ¶133). This requires a feature analyzer (i.e., processor per p. 17, ¶166, processor 166) applying either a Vision Transformer model to at least obtain / identify the entity certain athlete as the first entity or a BERT model to obtain / identify the first entity in the text modality (p. 6, ¶56).
Furthermore, the processor applied the Vision Transformer model to obtain / identify the first entity’s unimodal relationship in the image modality and applied the BERT model to obtain / identify the first entity’s unimodal relationship in the text modality (p. 6, ¶56) in order to obtain entities and relationships between entities from pictures, texts, and literature materials to construct the knowledge graph (p. 4, ¶40).
By obtaining / extracting at least the first entity (i.e., athlete) and first entity unimodal relationship in the text modality, the processor uses the BERT transformer model (p. 6, ¶56) to map the first entity (i.e., athlete) to a respective first entity – relationship – second entity entry in the knowledge graph.
Therefore, Zhao teaches “map, using the first ML model trained as the mapping agent, each of the one or more entities to respective one or more entries in a knowledge base to provide one or more entity mappings”.
In response to “The output described in Zhao is a prediction result, i.e., the second entity that completes the triple of the knowledge graph. (See paragraph 95 on pages 9 and 10, paragraph 134 on page 13, and paragraph 149 on page 15 of Zhao.) The embedding scores described in Zhao are intermediate values that are consumed by the weighted summation used to select that second entity, and Zhao does not describe providing any embedding score as an output” and “Consequently, Applicant respectfully submits that Zhao also fails to disclose the limitation: "provide an output identifving the content, at least one of the one or more entity mappings and the relevance score for the at least one of the one or more entity mappings,"”.
Zhao teaches a structure where module 601 acquires the first entity and the unimodal relationship of the first entity in multiple modalities (p. 15, ¶146), module 602 predicts the candidate entity corresponding to the first entity based on the first entity in each modality and unimodal relationship of the first entity (p. 15, ¶147), module 603 calculates embedding score of the candidate entity based on the first entity, the unimodal relationship of the first entity, and the candidate entity (p. 15, ¶148; per p. 7, ¶61, the embedding score characterizes the probability of the candidate entity being the second entity in the first entity-relationship-second entity in the knowledge graph triple; i.e., embedding score is a relevance score indicating how relevant the candidate entity is to the first entity and the unimodal relationship with the first entity so as to become the second entity in a corresponding knowledge graph triple), and module 604 determines the second entity from the candidate entity based on the embedding scores under each modality and weights of each modality (p. 15, ¶149).
In the example of Fig. 5, where entity is an athlete, the unimodal relationship is nationality, and the missing entity is correctly predicted to be country A (p. 13, ¶134) such that athlete – nationality – country A are the results of modules 601-603 outputting the triple athlete – nationality – country A and corresponding embedding scores in the structural, text, and image modalities to module 604 to make the final determination.
Therefore, Zhao teaches “provide an output identifying the content, at least one of the one or more entity mappings and the relevance score for the at least one of the one or more entity mappings”.
In response to “What Zhao combines are numerical scores obtained for a single candidate entity under the respective modalities, and nowhere does Zhao disclose aggregating entity mappings, identifying a set of aggregated entity mappings referencing a same entity, or providing an output that identifies any such set. (See generally Zhao.) Thus, Applicant respectfully submits that claims 5 and 15 are patentably distinguishable over Zhao for this additional reason”.
Zhao discloses obtaining entities and the relationships between entities from a large amount of pictures, texts, and other literature materials (p. 4, ¶40) where text modality is used to represent entity text information in the knowledge graph and image modality is used to represent entity image information in the knowledge graph (p. 5, ¶52).
Further, Zhao uses the TransE model to extract the first entity and the first entity’s unimodal relationship in the structural modality, uses the Vision Transformer model to extract the first entity and the first entity’s unimodal relationship in the image modality, and uses the BERT model to extract the first entity and the first entity’s unimodal relationship in the text modality (p. 6, ¶56).
Here, each extraction corresponds to a first entity – relationship mapping of the knowledge graph in the respective modality referencing the same first entity such that module 603 obtains an embedding score of at least one candidate entity in a corresponding modality based on the respective unimodal relationship of the first entity, the first entity, and the candidate entity (p. 15, ¶148; per p. 7, ¶61, the embedding score characterizes the probability of the candidate entity being the second entity in the first entity-relationship-second entity in the knowledge graph triple; i.e., each embedding score indicates the relevance of the candidate entity in the first entity – relationship – candidate entity mapping in the knowledge graph of a respective modality).
The module 604 determines the second entity for the first entity, and the unimodal relationship of the first entity from the candidate entity (p. 15, ¶149) by using a trained weight model (i.e., third ML model) to determine weights for respective modality (p. 16, ¶157) to perform summation (i.e., aggregation) of embedding scores of respective first entity- relationship – candidate entity mapping in the knowledge graph (p. 16, ¶156; see in particular, p. 10, ¶98, embedding score combination according to equation (7):
PNG
media_image2.png
33
334
media_image2.png
Greyscale
).
Therefore, the aggregated / combination embedding score corresponds to aggregated entity mappings referencing the same entity (first entity – second entity) because the individual embedding score before aggregation indicates the probability of a particular first entity – relationship – candidate entity as second entity mapping in the knowledge graph.
Finally, the determination of the second entity from the candidate entity is based on the output of the weighted aggregation / summation of embedding scores of respective first entity- relationship – candidate entity mappings (p. 16, ¶156).
Therefore, Zhao teaches “before the determining, aggregate, using a third ML model trained as an aggregation agent, all entity mappings of the plurality of entity mappings referencing a same entity of the plurality of entities to identify a set of aggregated entity mappings referencing the same entity” and “wherein the output further identifies the set of aggregated entity mappings”.
In response to “Paragraph 158 on page 16 of Zhao, relied upon in the Office Action, describes how the weight model of Zhao is trained and identifies no model type. Zhao describes that weight model as a multilayer perceptron using a hidden layer as a meta learner, and describes the only other weightings as an equal weighting, which involves no trained model, and a weighting generated by the RankBoost algorithm. (See paragraphs 102, 103 and 106 on page 10, and paragraph 125 on page 12 of Zhao.) None of those is an LLM or a multimodal foundation model, and the BERT model and the ViT model of Zhao serve as encoders of acquisition module 601, rather than to perform any aggregation. (See id. at paragraph 151 on page 15 and Figure 4.) Thus, Applicant respectfully submits that claims 6 and 16 are patentably distinguishable over Zhao for this additional reason”.
Zhao at Fig. 4 shows at least a first multimodal foundation model (i.e., an encoder comprising TransE model, BERT model, and ViT model) and a second multimodal foundation model (p. 7, ¶66, decoder calculates embedding score through score function fm (h, r, t) = f (hm, rm, tm), formula (1), where m -> M = {S, V, T} for structural, visual, and text modalities):
PNG
media_image3.png
366
607
media_image3.png
Greyscale
Since the decoder uses the trained weight model (p. 16, ¶157) to compute the combination embedding score (p. 10, ¶99), the decoder is equivalent of the third multi-modal foundation model.
Liu et al. (US 2025/0200293 A1) noted that such encoder-decoder structure comprises a LLM designed to process and understand multimodal data including audio, text, image, and video to perform some task (¶30 and ¶32).
Claim Rejections - 35 USC § 103
The following is a quotation of the appropriate paragraphs of 35 U.S.C. 103 that form the basis for the rejections under this section made 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, 4-5, 7-8, 10-11, 14-15, 17-18, and 20 are rejected under 35 USC 103 for being unpatentable over Zhao et al. (WO 2024/120385 A1; see attached translation) in view of Li et al. (US 9256682 B1).
Regarding Claims 1 and 11, Zhao discloses a system (p. 5, ¶46 and Figs. 6-7) comprising:
a computing platform including a hardware processor and a system memory (p. 5, ¶46, server or terminal device implemented by a processor calling a computer readable program instruction stored in a memory);
an entity tracing engine stored in the system memory, the entity tracing engine including a first machine learning (ML) model trained as a mapping agent and a second ML model trained as a scoring agent (Fig. 6, device 600 includes an acquisition module 601 using TransE model, ViT model, or a BERT model per p. 15, ¶151 and prediction module 602 using pre-trained prediction model per p. 15, ¶152);
the hardware processor configured to execute the entity tracing engine to:
receive content including at least one of an image, a video, an audio, or a text (Fig. 4, receiving at least text data and image data);
identify, using a feature analyzer, one or more entities referenced in the content (p. 5, ¶63, use BERT model to extract first entity; p. 5, ¶64, use ViT model to extract first entity);
map, using the first ML model trained as the mapping agent, each of the one or more entities to respective one or more entries in a knowledge base to provide one or more entity mappings (p. 7, ¶64, image data uses ViT model to (1) extract the first entity and unimodal relationship with first entity, and (2) obtain relationship between the entity with image features and image modality; p. 7, ¶63, text data uses BERT model to (1) extract first entity and unimodal relationship with the first entity, and (2) obtain relationship between the entity with text features and the text modality; per p. 1, ¶10; i.e., an entry in the knowledge graph comprises two entities and an edge relationship or triple such that extracting first entity and first entity unimodal relationship in the respective modality maps the first entity to a corresponding entry in the knowledge graph);
determine, using the second ML model trained as the scoring agent, a relevance score for each of the one or more entity mappings relative to the content (p. 6, ¶55 and p. 8, ¶71, different forms of decoders use different models where text modality trains a text prediction model and image modality trains an image prediction model; pp. 6-7, ¶60, decoding is performed based on single modal relationship to obtain embedding score of each modality; p. 7, ¶61, predict at least one candidate entity corresponding to the first entity based on the first entity in each modality and the unimodal relationship of the first entity and obtain an embedding score of the at least one candidate entity); and
provide an output identifying the content, at least one of the one or more entity mappings and the relevance score for the at least one of the one or more entity mappings (pp. 10-11, ¶109, obtain the final prediction result in combination with the modal score; p. 15, ¶149, determination module 604 determines second entity from at least one candidate entity based on the embedding scores to obtain triples in the knowledge graph comprising the first entity, the unimodal relationship, and the second entity; e.g., p. 13, ¶¶133-34, based on the text, structural, and image, determine that the entity is an athlete, the unimodal relationship is nationality, and predict the second entity to be country A by increasing the weight of the text modality and reducing the weight of image modality).
Zhao does not disclose the relevance score is determined based on a number of times a respective one of the one or more entities is referenced in the content.
Li discloses a processor configured to identify at least one entity and related entities related to the at least one entity by respective properties in a knowledge graph (Col 1, Rows 39-41) and for each respective one of the related entities, determine a related entity score associated with a respective property that relates the at least one entity and the respective one of the related entities (Col 1, Rows 41-45), wherein the related entity score indicates a number of times a respective one of the related entities is referenced in a content (Col 9, Rows 49-53, each pair of the entity and a related entity may have an associated related entity score indicating the relevance and instances of co-occurrence on one or more webpages).
Zhao discloses the embedding score characterizes the probability that a candidate entity is the second entity in a knowledge graph triple comprising the first entity – relationship – second entity (p. 7, ¶61); i.e., embedding score is the probability that the candidate entity relates to the first entity by the corresponding relationship.
It would’ve been obvious to one ordinarily skilled in the art before the effective filing date of the invention to calculate the embedding score / related entity score based on a number of times a respective one of the one or more entities is referenced in the content (i.e., the co-occurrence of the first entity and the second entity on one or more content per Zhao p. 4, ¶40) in order to determine the relevance of related entities (Li, Col 9, Rows 49-51; compare Zhao, p. 7, ¶61) for a data structure of sortable properties usable to provide sorted search results in response to a query (Li, Col 1, Rows 33-35).
Regarding Claims 4 and 14, Zhao discloses wherein the hardware processor is further configured to execute the entity tracing engine to:
identify, based on the content, a context for tracing the one or more entities (p. 16, ¶156, determination module 604 perform weighted summation on each candidate entity based on the weights of each modality and the embedding scores of the candidate entities under each modality to obtain a weighted summation result of each candidate entity);
wherein each of the mapping and the determining uses the context (p. 13, ¶133, Fig. 5, text modality introduces entity (a certain athlete), structural modality contains athlete’s unimodal relationships, and image modality shows the scene of the athlete playing where text modality is helpful, image modality is of no help, and structural modality has both useful and useless information; p. 13, ¶134, adjust the weights of the text modality, image modality, and structural modality such as increasing the weight of the text modality and reducing the weight of the image modality to improve prediction accuracy where based on entity is an athlete and unimodal relationship is nationality in the text modality, predict the second entity as country A; i.e., p. 15, ¶149, weights characterize the importance of each modality being the context, which are used to improve prediction accuracy).
Regarding Claims 5 and 15, Zhao discloses wherein the one or more entities include a plurality of entities, the one or more entity mappings include a plurality of entity mappings (p. 7, ¶64, image data uses ViT model to (1) extract the first entity and unimodal relationship with first entity, and (2) obtain relationship between the entity with image features and image modality; p. 7, ¶63, text data uses BERT model to (1) extract first entity and unimodal relationship with the first entity, and (2) obtain relationship between the entity with text features and the text modality), and wherein the hardware processor is further configured to execute the entity tracing engine to:
before the determining
aggregate, using a third ML model trained as an aggregation agent (p. 16, ¶157, weights under each modality are determined based on a trained weight model), all entity mappings of the plurality of entity mappings referencing a same entity of the plurality of entities to identify a set of aggregated entity mappings referencing the same entity (p. 16, ¶156, determination module 604 perform weighted summation on each candidate entity based on the weights of each modality and the embedding scores of the candidate entities under each modality to obtain a weighted summation result of each candidate entity);
wherein the output further identifies the set of aggregated entity mappings (p. 16, ¶156, perform weighted summation on each candidate entity based on the weights of each modality and the embedding scores of the candidate entities under each modality to obtain a weighted summation result of each candidate entity and determine the second entity corresponding to the first entity based on the weighted summation result of each candidate entity).
Regarding Claims 7 and 17, Zhao discloses wherein each of the one or more entity mappings includes an identity of an entity mapped by the entity mapping (p. 13, ¶133, text modality introduces the entity as a certain athlete; p. 13, ¶134, in the exemplary entity-relationship-entity mapping of Fig. 5, the identity of the missing entity was correctly predicted to be country A), an entity type of the entity (p. 13, ¶134, in the exemplary entity-relationship-entity mapping of Fig. 5, the type of the missing entity is nationality according to the unimodal relationship; per p. 1, ¶10, nodes and edges are used to represent triples consisting of entity-relationship (or attribute)-entity), and a knowledge base address of a knowledge base entry referencing the entity (p. 13, ¶134, complete the knowledge graph based on the triplet athlete-nationality-country A; i.e., p. 4, ¶40, knowledge graph / base are constructed (i.e., addressed with entries or nodes per p. 1, ¶10 where nodes represent entities and edge between two nodes represent relationship between two entities) based on entities and relationship triples).
Regarding Claims 8 and 18, Zhao discloses wherein the feature analyzer (Fig. 4 showing an encoder and a decoder, the encoder comprising Vision Transformer for image data, BERT model for text, and TransE model for structural data) includes at least one of a facial recognition module (Fig. 4, Vision Transformer in the encoder analyzes image data to predict entity ev; Fig. 5 and p. 13, ¶133, Vision Transformer analyzes image data showing scene of the athlete playing), an object recognition module, an activity recognition module (Fig. 4, Vision Transformer in the encoder analyzes image data to predict entity ev; Fig. 5 and p. 13, ¶133, Vision Transformer analyzes image data showing scene of the athlete playing), or a text analysis module (Fig. 4, BERT in the encoder analyzes text data to predict entity et; p. 13, ¶133, BERT analyzes text data introducing the entity as a certain athlete) configured to analyze text and speech included in the content (p. 14, ¶140, different decoders are used to obtain candidate entities in text and audio modality and determine the second entity based on the text embedding score and the audio embedding score).
Regarding Claims 10 and 20, Zhao discloses wherein the content comprises at least one of sports content (p. 13, ¶133, text modality introduces the entity as a certain athlete while image modality shows the scene of the athlete playing), television programming content, movie content, advertising content, or video game content.
Claims 2-3, 6, 12-13, and 16 are rejected under 35 USC 103 for being unpatentable over Zhao et al. (WO 2024/120385 A1; see attached translation) and Li et al. (US 9256682 B1) as applied to claims 1 and 11, in view of Liu et al. (US 2025/0200293 A1).
Regarding Claims 2 and 12, Zhao discloses wherein the mapping agent is implemented using a first large-language model (LLM) or a first multimodal foundation model (p. 6, ¶56, p. 7, ¶63, and p. 15, ¶151, acquisition module 601 comprising TransE model for structural modality, ViT (Vision Transformer) model in image modality, and BERT model for text modality; Fig. 4 showing an encoder comprising the TransE model, BERT model, and ViT model), and wherein the scoring agent is implemented using a second LLM or a second multimodal foundation model (p. 7, ¶62 and p. 7, ¶71, Fig. 4, decoder comprising prediction models for respective structural modality, text modality, and image modality).
Zhao does not disclose implementing the mapping agent (i.e., encoder) and the scoring agent (i.e., decoder) using LLMs.
Liu discloses multi-modal LLMs to process and understand audio, text, image, and video (¶30) comprising an encoder-decoder structure where encoder produces a representation of text and image and where the decoder uses the representation to perform a task (¶32).
It would’ve been obvious to one ordinarily skilled in the art before the effective filing date of the invention to implement at least a first LLM with encoder to process text and image (i.e., Zhao, p. 7, ¶¶63-64, Fig. 4, encoder with BERT model and Vision Transformer model; compare Liu, ¶69, LLM with BERT an image generation model) to implement the mapping agent / encoder and at least a second LLM with decoder to perform a task (i.e., Zhao, p. 7, ¶¶64-66 and Fig. 4, decoder calculates embedding score) in order to process and understand at least text and image (Liu, ¶30).
Regarding Claims 6 and 16, Zhao discloses wherein the aggregation agent is implemented using a third LLM or a third multimodal foundation model (p. 16, ¶158, training step of the weight model includes an inner loop and an outer loop, the inner loop is a loop between multiple modalities).
Zhao does not disclose implementing the aggregation agent as a third LLM.
Liu discloses multi-modal LLMs to process and understand audio, text, image, and video (¶30) comprising an encoder-decoder structure where encoder produces a representation of text and image and where the decoder uses the representation to perform a task (¶32).
It would’ve been obvious to one ordinarily skilled in the art before the effective filing date of the invention to implement the decoder aggregation agent (Zhao p. 10, ¶¶98-99 and p. 16, ¶¶156-57, Fig. 4, decoder using weighted model to generate summation embedding score of first entity – relationship – candidate entity mapping in respective modalities according to weights of respective modality) with at least a third LLM with decoder to perform a task (i.e., Zhao, p. 7, ¶¶64-66 and Fig. 4, decoder calculates embedding score) in order to process and understand at least text and image (Liu, ¶30).
Regarding Claims 3 and 13, Zhao does not disclose wherein at least the first LLM or the first multimodal foundation model is configured to perform one or more of zero-shot learning or few-shot learning.
Liu discloses configuring LLM to perform one or more of zero-shot learning or few-shot learning (¶37, the language model (i.e., LLM) configured using zero-shot or few-shot capability).
It would’ve been obvious to one ordinarily skilled in the art before the effective filing date of the invention to implement the first multimodal foundation model in Liu’s encoder (Fig. 4) to perform zero-shot learning or few-shot learning in order to implement the first LLM that requires fewer amount of training data (Liu, ¶37).
Claims 9 and 19 are rejected under 35 USC 103 for being unpatentable over Zhao et al. (WO 2024/120385 A1; see attached translation) and Li et al. (US 9256682 B1) as applied to claims 1 and 11, in view of Farre Guiu et al. (US 11354894 B2).
Regarding Claims 9 and 19, Zhao does not disclose wherein the feature analyzer includes at least one of an organization recognition module or a venue recognition module.
Farre Guiu discloses a feature analyzer to apply labels to features detected in the content, access one or more knowledge bases to validate the applied labels, and obtain descriptive data linked to the validated labels from the knowledge base (Abstract and Col 2, Rows 3-14), the feature analyzer includes an organization recognition module or a venue recognition module (Col 5, Rows 25-31).
It would’ve been obvious to one ordinarily skilled in the art before the effective filing date of the invention to implement the feature analyzer / encoder of Zhao (Fig. 4) to include an organization recognition module or a venue recognition module to perform text analysis of the content to identify venue, brands, logo or organization of the content (Farre Guiu, Col 5, Rows 25-31).
Conclusion
Applicant's amendment necessitated the new grounds 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 mailing date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to examiner Richard Z. Zhu whose telephone number is 571-270-1587 or examiner’s supervisor Hai Phan whose telephone number is 571-272-6338. Examiner Richard Zhu can normally be reached on M-Th, 0730:1700.
Information regarding the status of an application may be obtained from the Patent Application Information Retrieval (PAIR) system. Status information for published applications may be obtained from either Private PAIR or Public PAIR. Status information for unpublished applications is available through Private PAIR only. For more information about the PAIR system, see http://pair-direct.uspto.gov. Should you have questions on access to the Private PAIR system, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative or access to the automated information system, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/RICHARD Z ZHU/Primary Examiner, Art Unit 2654 10/14/2026
