DETAILED ACTION
Notice of Pre-AIA or AIA Status
1. The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Drawings
2. The drawings filed on 05/28/2025 are accepted by the examiner.
Information Disclosure Statement
3. The information disclosure statement (IDS) submitted on 28 May 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
4. Applicant’s amendments to the Specification filed on May 28, 2025, has been considered and entered.
Claim Objections
5. Claim 9 is objected to because of the following informalities:
In claim 9 line 4, “the processor” should read “the at least one processor”.
Appropriate correction is required.
Claim Rejections - 35 USC § 112
The following is a quotation of 35 U.S.C. 112(b):
(b) CONCLUSION.—The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the inventor or a joint inventor regards as the invention.
The following is a quotation of 35 U.S.C. 112 (pre-AIA ), second paragraph:
The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the applicant regards as his invention.
6. Claims 1-8, 11, 13-16 and 19 are rejected under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), second paragraph, as being indefinite for failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor, or for pre-AIA the applicant regards as the invention.
Claim 1 recites the limitations "the computer system" in line 2 and “the at least one processor” in line 4. There is insufficient antecedent basis for these limitations in the claim.
Claims 3, 11 and 19 recite the limitations "the scoring of the similarity" and “the plurality of vectors”. There is insufficient antecedent basis for these limitations in the claims.
Claims 4 and 8 recite the limitation "the at least one processor" in line 2. There is insufficient antecedent basis for this limitation in the claims.
Claims 5, 7-8, 13 and 15-16 recite the limitation "the plurality of vectors". There is insufficient antecedent basis for this limitation in the claims.
The remaining claims are rejected for fully incorporating the deficiencies of the base claim(s) from which they depend.
Claim Rejections - 35 USC § 102
7. 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 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 the appropriate paragraphs of 35 U.S.C. 102 that form the basis for the rejections under this section made in this Office action:
A person shall be entitled to a patent unless –
(a)(2) the claimed invention was described in a patent issued under section 151, or in an application for patent published or deemed published under section 122(b), in which the patent or application, as the case may be, names another inventor and was effectively filed before the effective filing date of the claimed invention.
8. Claims 1-20 are rejected under 35 U.S.C. 102(a)(2) as being anticipated by KOPRU et al. (US 2023/0101174 A1) hereinafter KOPRU.
As to claim 1, KOPRU discloses a computer-implemented method comprising: receiving, at the computer system (Para. 70), a query record from a user (Fig. 6, Para. 29, The query receiver 134 receives a query from the buyer, i.e., a user, using the client device 102, the computer system, via the network 160. In aspects, the query receiver transforms the query (e.g., in a text form) into embedding vector data 150 by using the trained model. In aspects, the embedding vector data 150 describes similarity distance data in multi-dimensional space.), the query record comprising a plurality of attributes (Fig. 6, Para. 56, “the receive operation 604 receives a query from a buyer using a client device (e.g., the client device 102 as shown in FIG. 1) and access the online shopping system. The query may include a command and one or more parameters needed to search for items and products.”. Para. 57, “The embedded vector data includes a multidimensional vector including attributes and values, which in aggregate represent features sought during a search. For example, attributes may include items, products, users, image data, and the like.”. Thus, the query record comprising a plurality of attributes.);
executing, via the at least one processor (Para. 70), a trained siamese neural network using the query record from the user as input, resulting in a vector representation of the query record from the user (Fig. 5, Para. 21, “a Siamese neural network setting may be used to train the transformer model that generates the vector embeddings.”. Para. 52, “the train operation 508 includes training a Siamese neural network, which uses the same weight for processing two input vectors in tandem and generates an output vector.”. Para. 70, “generating, based on the received query, embedded vector data using a model, wherein the embedded vector data indicates vector representations of similarities among the received query and items. The model includes a Siamese network, and the computer-executable instructions when executed further causing the system to execute a method further comprising retrieving a pair of input from the training data, the pair of input indicating ground truth examples of one or more of: item-to-item, user-to-item, product-to-item, or query-to-item; and training, based on the pair of input, the Siamese network, the Siamese network including a plurality of encoders, each encoder encoding one of the pair of input to generate embedding vector data.”. Thus, executing, via the at least one processor, a trained siamese neural network using the query record from the user as input, resulting in a vector representation of the query record from the user.);
executing, via the at least one processor using the vector representation, a search of a plurality of records, resulting in similar records (Fig. 3, Para. 40, “The query receiver & listing retriever 212 generates embedded vectors using the trained Transformer model and stores online embedded vector data 214. The kNN searcher 216 uses the combination of the online embedded vector data 214 associated with the received query and kNNs in the kNN index data 210 and generates a listing of items.”. Para. 41, “FIG. 3 illustrates an example of a received query and listings of items as search results according to aspects of the present disclosure. The diagram 300 includes a search 302 and a result without similarity 304 and a result with similarity 306. In aspects, the latter list, the result with similarity 306 describes a result based on the similarity index according to the present disclosure”. Para. 70, “generating, based on the received query, embedded vector data using a model, wherein the embedded vector data indicates vector representations of similarities among the received query and items, and wherein the model includes a trained model based on similarities at least in one or more relationships including: item-to-item, user-to-item, or query-to- item; determining, based on a similarity index search using similarity index data and the embedded vector data, one or more items for listing.”. Thus, executing, via the at least one processor using the vector representation, a search of a plurality of records, resulting in similar records.); and
providing the similar records to the user in response to the query record from the user (Para. 30, “The listing generator 136 generates a listing of items (e.g., a listing of products) as a search result of the received query.”. Para. 40, The listing generator generates a graphical representation of the listing of items and transmits the listing to the client device for display to the user, i.e., providing the similar records to the user in response to the query record from the user.).
As to claim 9, KOPRU discloses a system comprising: at least one processor; and a non-transitory computer-readable storage medium storing instructions which. when executed by the processor (Para. 70), cause the at least one processor to perform operations comprising:
receiving a query record from a user (Fig. 6, Para. 29, The query receiver 134 receives a query from the buyer, i.e., a user, using the client device 102, the computer system, via the network 160. In aspects, the query receiver transforms the query (e.g., in a text form) into embedding vector data 150 by using the trained model. In aspects, the embedding vector data 150 describes similarity distance data in multi-dimensional space.), a query record comprising a plurality of attributes (Fig. 6, Para. 56, “the receive operation 604 receives a query from a buyer using a client device (e.g., the client device 102 as shown in FIG. 1) and access the online shopping system. The query may include a command and one or more parameters needed to search for items and products.”. Para. 57, “The embedded vector data includes a multidimensional vector including attributes and values, which in aggregate represent features sought during a search. For example, attributes may include items, products, users, image data, and the like.”. Thus, a query record comprising a plurality of attributes.);
executing a trained siamese neural network using the query record from the user as input, resulting in a vector representation of the query record from the user (Fig. 5, Para. 21, “a Siamese neural network setting may be used to train the transformer model that generates the vector embeddings.”. Para. 52, “the train operation 508 includes training a Siamese neural network, which uses the same weight for processing two input vectors in tandem and generates an output vector.”. Para. 70, “generating, based on the received query, embedded vector data using a model, wherein the embedded vector data indicates vector representations of similarities among the received query and items. The model includes a Siamese network, and the computer-executable instructions when executed further causing the system to execute a method further comprising retrieving a pair of input from the training data, the pair of input indicating ground truth examples of one or more of: item-to-item, user-to-item, product-to-item, or query-to-item; and training, based on the pair of input, the Siamese network, the Siamese network including a plurality of encoders, each encoder encoding one of the pair of input to generate embedding vector data.”. Thus, executing a trained siamese neural network using the query record from the user as input, resulting in a vector representation of the query record from the user.);
executing, using the vector representation, a search of a plurality of records, resulting in similar records (Fig. 3, Para. 40, “The query receiver & listing retriever 212 generates embedded vectors using the trained Transformer model and stores online embedded vector data 214. The kNN searcher 216 uses the combination of the online embedded vector data 214 associated with the received query and kNNs in the kNN index data 210 and generates a listing of items.”. Para. 41, “FIG. 3 illustrates an example of a received query and listings of items as search results according to aspects of the present disclosure. The diagram 300 includes a search 302 and a result without similarity 304 and a result with similarity 306. In aspects, the latter list, the result with similarity 306 describes a result based on the similarity index according to the present disclosure”. Para. 70, “generating, based on the received query, embedded vector data using a model, wherein the embedded vector data indicates vector representations of similarities among the received query and items, and wherein the model includes a trained model based on similarities at least in one or more relationships including: item-to-item, user-to-item, or query-to- item; determining, based on a similarity index search using similarity index data and the embedded vector data, one or more items for listing.”. Thus, executing, using the vector representation, a search of a plurality of records, resulting in similar records.); and
providing the similar records to the user in response to the query record from the user (Para. 30, “The listing generator 136 generates a listing of items (e.g., a listing of products) as a search result of the received query.”. Para. 40, The listing generator generates a graphical representation of the listing of items and transmits the listing to the client device for display to the user, i.e., providing the similar records to the user in response to the query record from the user.).
As to claim 17, KOPRU discloses a non-transitory computer-readable storage medium storing instructions which, when executed by at least one processor (Para. 70), cause the at least one processor to perform operations comprising:
receiving a query record from a user (Fig. 6, Para. 29, The query receiver 134 receives a query from the buyer, i.e., a user, using the client device 102, the computer system, via the network 160. In aspects, the query receiver transforms the query (e.g., in a text form) into embedding vector data 150 by using the trained model. In aspects, the embedding vector data 150 describes similarity distance data in multi-dimensional space.), the query record comprising a plurality of attributes (Fig. 6, Para. 56, “the receive operation 604 receives a query from a buyer using a client device (e.g., the client device 102 as shown in FIG. 1) and access the online shopping system. The query may include a command and one or more parameters needed to search for items and products.”. Para. 57, “The embedded vector data includes a multidimensional vector including attributes and values, which in aggregate represent features sought during a search. For example, attributes may include items, products, users, image data, and the like.”. Thus, the query record comprising a plurality of attributes.);
executing a trained siamese neural network using the query record from the user as input, resulting in a vector representation of the query record from the user (Fig. 5, Para. 21, “a Siamese neural network setting may be used to train the transformer model that generates the vector embeddings.”. Para. 52, “the train operation 508 includes training a Siamese neural network, which uses the same weight for processing two input vectors in tandem and generates an output vector.”. Para. 70, “generating, based on the received query, embedded vector data using a model, wherein the embedded vector data indicates vector representations of similarities among the received query and items. The model includes a Siamese network, and the computer-executable instructions when executed further causing the system to execute a method further comprising retrieving a pair of input from the training data, the pair of input indicating ground truth examples of one or more of: item-to-item, user-to-item, product-to-item, or query-to-item; and training, based on the pair of input, the Siamese network, the Siamese network including a plurality of encoders, each encoder encoding one of the pair of input to generate embedding vector data.”. Thus, executing a trained siamese neural network using the query record from the user as input, resulting in a vector representation of the query record from the user.);
executing, using the vector representation, a search of a plurality of records, resulting in similar records (Fig. 3, Para. 40, “The query receiver & listing retriever 212 generates embedded vectors using the trained Transformer model and stores online embedded vector data 214. The kNN searcher 216 uses the combination of the online embedded vector data 214 associated with the received query and kNNs in the kNN index data 210 and generates a listing of items.”. Para. 41, “FIG. 3 illustrates an example of a received query and listings of items as search results according to aspects of the present disclosure. The diagram 300 includes a search 302 and a result without similarity 304 and a result with similarity 306. In aspects, the latter list, the result with similarity 306 describes a result based on the similarity index according to the present disclosure”. Para. 70, “generating, based on the received query, embedded vector data using a model, wherein the embedded vector data indicates vector representations of similarities among the received query and items, and wherein the model includes a trained model based on similarities at least in one or more relationships including: item-to-item, user-to-item, or query-to- item; determining, based on a similarity index search using similarity index data and the embedded vector data, one or more items for listing.”. Thus, executing, using the vector representation, a search of a plurality of records, resulting in similar records.); and
providing the similar records to the user in response to the query record from the user (Para. 30, “The listing generator 136 generates a listing of items (e.g., a listing of products) as a search result of the received query.”. Para. 40, The listing generator generates a graphical representation of the listing of items and transmits the listing to the client device for display to the user, i.e., providing the similar records to the user in response to the query record from the user.).
As to claims 2, 10 and 18, the claims are rejected for the same reasons as claims 1, 9 and 17 above. In addition, KOPRU discloses wherein the plurality of records are K nearest neighbors of the query record, K being a predetermined number, and wherein the plurality of records are identified as nearest neighbors based on vector similarity (Fig. 5, Para. 22, “The disclosed technology generates a similarity index based on the embedded vector representation of data. In particular, a k-Nearest Neighborhood (kNN) index generator generates a data structure that represents kNN indices. For example, the disclosed technology may store the kNN index data in a hierarchical graph with a plurality of layers of nodes (e.g., Hierarchical Navigable Small World (HNSW) graph). The disclosed technology does not limit itself to use HNSW graph. Other examples may include k-selection methods in a nearest neighborhood in scalable settings and in compressed vector settings.”. Para. 30, “The listing generator 136 generates a listing of items (e.g., a listing of products) as a search result of the received query. In aspects, the listing generator 136 determines a set of items with similarity within a k-nearest neighborhood in the kNN index data 152.”. Thus, the plurality of records are K nearest neighbors of the query record, K being a predetermined number, and wherein the plurality of records are identified as nearest neighbors based on vector similarity.).
As to claims 3, 11 and 19, the claims are rejected for the same reasons as claims 1, 9 and 17 above. In addition, KOPRU discloses wherein the scoring of the similarity between records is performed using weights associated with each of the plurality of vectors representing attributes, values of the weights representing relative importance of each attribute (Para. 52, “the train operation 508 includes training a Siamese neural network, which uses the same weight for processing two input vectors in tandem and generates an output vector. For example, the two input vectors may represent a true example pair of an item and another item, an item and an image associated with the item, a user (i.e., a history of user interaction in searching and selecting items as a past event) and an item, a query and an item, and the like.”. Para. 33, “The kNN index updater 142 updates the kNN index data 152 when a need arises to modify the similarities among items, products, queries, and users as the buyer operates on the listing of items. For example, the kNN index updater 142 may update an index value associated with the item when the buyer selects the item on the list of items for acquisition, signifying that the selected item has a preference that is higher than other items on the list.”. Para. 57, “The embedded vector data includes a multidimensional vector including attributes and values, which in aggregate represent features sought during a search. For example, attributes may include items, products, users, image data, and the like.”. Thus, the scoring of the similarity between records is performed using weights associated with each of the plurality of vectors representing attributes, values of the weights representing relative importance of each attribute.).
As to claims 4, 12 and 20, the claims are rejected for the same reasons as claims 1, 9 and 17 above. In addition, KOPRU discloses further comprising: executing, via the at least one processor prior to the execution of the search of the plurality of records, the siamese neural network using the plurality of records as input, resulting in a plurality of vector representations corresponding to individual records within the plurality of records, wherein the search of the plurality of records comprises a comparison of the vector representation to the plurality of vector representations (Fig. 5, Para. 52, “the train operation 508 includes training a Siamese neural network, which uses the same weight for processing two input vectors in tandem and generates an output vector. For example, the two input vectors may represent a true example pair of an item and another item, an item and an image associated with the item, a user (i.e., a history of user interaction in searching and selecting items as a past event) and an item, a query and an item, and the like.”. Para. 70, “generating, based on the received query, embedded vector data using a model, wherein the embedded vector data indicates vector representations of similarities among the received query and items. The model includes a Siamese network, and the computer-executable instructions when executed further causing the system to execute a method further comprising retrieving a pair of input from the training data, the pair of input indicating ground truth examples of one or more of: item-to-item, user-to-item, product-to-item, or query-to-item; and training, based on the pair of input, the Siamese network, the Siamese network including a plurality of encoders, each encoder encoding one of the pair of input to generate embedding vector data.”. Thus, the search of the plurality of records comprises a comparison of the vector representation to the plurality of vector representations.).
As to claims 5 and 13, the claims are rejected for the same reasons as claims 1 and 9 above. In addition, KOPRU discloses wherein the plurality of vectors representing attributes are generated using Natural Language Processing (Para. 4, “the above and other issues are resolved by automatically generating similarity index data using a machine learning model and using the similarity index data to determine similarities among items, products, users, and queries”. Para. 5, “the present disclosure uses artificial intelligence that is robust, scalable, and accurate in identifying goods, queries, buyers, sellers, messages among buyers and sellers, product reviews, user reviews, and the like, which are similar to one another. Artificial intelligence includes one or more trained models (e.g., a transformer model, a neural network model, and the like) for predicting similarities.”. Para. 29, “The query receiver 134 receives a query from the buyer using the client device 102 via the network 160. In aspects, the query receiver transforms the query (e.g., in a text form) into embedding vector data 150 by using the trained model.”. Thus, the plurality of vectors representing attributes are generated using Natural Language Processing.).
As to claims 6 and 14, the claims are rejected for the same reasons as claims 5 and 13 above. In addition, KOPRU discloses wherein the Natural Language Processing comprises at least one of (A) a neural network model and (B) a transformer-based machine learning algorithm (Para. 4, “the above and other issues are resolved by automatically generating similarity index data using a machine learning model and using the similarity index data to determine similarities among items, products, users, and queries.”. Para. 5, “the present disclosure uses artificial intelligence that is robust, scalable, and accurate in identifying goods, queries, buyers, sellers, messages among buyers and sellers, product reviews, user reviews, and the like, which are similar to one another. Artificial intelligence includes one or more trained models (e.g., a transformer model, a neural network model, and the like) for predicting similarities.”. Thus, the Natural Language Processing comprises at least one of (A) a neural network model and (B) a transformer-based machine learning algorithm.).
As to claims 7 and 15, the claims are rejected for the same reasons as claims 1 and 9 above. In addition, KOPRU discloses wherein the plurality of vectors representing attributes comprise: a unit identification; at least one unit classification; a receiving entity; and a geographic identification (Para. 5, “the present disclosure uses artificial intelligence that is robust, scalable, and accurate in identifying goods, queries, buyers, sellers, messages among buyers and sellers, product reviews, user reviews, and the like, which are similar to one another. Artificial intelligence includes one or more trained models (e.g., a transformer model, a neural network model, and the like) for predicting similarities.”. Para. 57, “The embedded vector data includes a multidimensional vector including attributes and values, which in aggregate represent features sought during a search. For example, attributes may include items, products, users, image data, and the like.”. Thus, the plurality of vectors representing attributes comprise: a unit identification; at least one unit classification; a receiving entity; and a geographic identification.).
As to claims 8 and 16, the claims are rejected for the same reasons as claims 1 and 9 above. In addition, KOPRU discloses further comprising: scoring, via the at least one processor of a computer system, similarity between records in a training set of data, each record within the training set of data comprising the plurality of vectors representing attributes, resulting in at least one similarity score (Fig. 6, Para. 4, the above and other issues are resolved by automatically generating similarity index data using a machine learning model and using the similarity index data to determine similarities among items, products, users, and queries, i.e., similarity between records in a training set of data. Para. 29, “The query receiver 134 receives a query from the buyer using the client device 102 via the network 160. In aspects, the query receiver transforms the query (e.g., in a text form) into embedding vector data 150 by using the trained model. In aspects, the embedding vector data 150 describes similarity distance data in multi-dimensional space.”.); and
training, via the at least one processor using the at least one similarity score and the training set of data, a siamese neural network, resulting in the trained siamese neural network, wherein the trained siamese neural network is configured to receive one or more records and the siamese neural network is configured to generate vector representations for the one or more records (Para. 21, “a Siamese neural network setting may be used to train the transformer model that generates the vector embeddings.”. Para. 44, “The difference in the listing may be based on the learnt data that is trained to determine similarities associated with items, products, queries, and buyers during the pre-training of the Transformer model.”. Para. 52, “the train operation 508 includes training a Siamese neural network, which uses the same weight for processing two input vectors in tandem and generates an output vector. For example, the two input vectors may represent a true example pair of an item and another item, an item and an image associated with the item, a user (i.e., a history of user interaction in searching and selecting items as a past event) and an item, a query and an item, and the like”. Para. 53, “The generate operation 510 merges the pair of embedding vectors into embedded vectors by various processing including but not limited to mapping into a multimodal space, enriching embeddings by training for predicting randomly masked name-value pairs.”. Thus, the trained siamese neural network is configured to receive one or more records and the siamese neural network is configured to generate vector representations for the one or more records.).
Conclusion
9. The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
Misiewicz et al. (US 2022/0138170 A1) teaches vector-based search result generation.
Strope et al. (US 2018/0240013 A1) teaches training and/or using a relevance model for information retrieval.
Ramanath et al. (US 2019/0251422 A1) teaches deep neural network architecture for search.
Liu et al. (US 2018/0246966 A1) teaches search and retrieval system that increases the efficiency of search and retrieval by using on a predicted computerized rendering period.
10. Any inquiry concerning this communication or earlier communications from the examiner should be directed to MOHAMMAD SOLAIMAN BHUYAN whose telephone number is (571)272-7843. The examiner can normally be reached on Monday - Friday 9:00am-5:00pm EST.
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, Charles Rones can be reached on 571-272-4085. The fax phone number for the organization where this application or proceeding is assigned is 571 -273-8300.
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.
/Mohammad S Bhuyan/Examiner, Art Unit 2168
/CHARLES RONES/Supervisory Patent Examiner, Art Unit 2168
