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 .
DETAILED ACTION
Summary
This action is in reply to Applicant’s Amendments and Remarks filed on 7/21/2026.
Claims 1-20 are pending.
Response to Arguments
Claims 1, 13 and 20 have been amended and are directed to eligible subject matter. Therefore, the rejections of claims 1-20 under 35 U.S.C. § 101 are withdrawn.
Claim 3, 6-12 and 18-19 have been amended and there is sufficient antecedent basis for the limitation in the claims. Therefore, the rejections of claims 3, 6-12 and 18-19 under 35 U.S.C. § 112(b) are withdrawn.
Applicant's arguments with respect to claim limitation “generating a second set of features for the second set of search queries based on the first set of features; and training a machine-learning search model based on the second set of search queries and the second set of features” recited in claim 1 have been fully considered but they are not persuasive. Applicants contend the combination of Oshio and LAMBA does not teach the claimed feature above (Applicants’ Remarks dated 7/21/2026, p. 5-7). However, the Examiner respectfully disagrees. Figures 6-7 of Oshio illustrate generating translated search query (in Japanese) including information of the language on search query and language information on search results based on the first search query [in English]; then re-training the machine-learning search model [at step 112 in FIG. 7] based on the translated search query [in Japanese] including information of the language on search query and language information on search results.
Applicant's arguments with respect to amended claims and originally presented claims have been fully considered but they are moot in view of the new grounds of rejection.
During patent examination, the pending claims must be "given their broadest reasonable interpretation consistent with the specification." Phillips v. AWH Corp., 415 F.3d 1303, at 1316 (Fed. Cir. 2005). See also In re Hyatt, 211 F.3d 1367, 1372, 54 USPQ2d 1664, 1667 (Fed. Cir. 2000).
Claim Objections
Claim 3 is objected to because of the following informalities: “the context for of” in the first row should be “the context . Appropriate correction is required.
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(s) 1, 13 and 20 is/are rejected under 35 U.S.C. 103 as being unpatentable over Oshio et al (US 20220335046 A1) in view of LAMBA et al (US 20230281399 A1) and Hwang et al (US 20150127319 A1).
Regarding claim 1, Oshio discloses a method [e.g. FIG. 2] comprising:
obtaining a first search log of a first set of search queries [e.g. FIG. 5-6; search queries; search query language information] in a first language [e.g. English; the second language];
generating one or more prompts for input [e.g. FIG. 7; [0064]; a search query may be prompted] to a machine-learned language model [e.g. FIG. 6-7; machine learning model], the one or more prompts including the first set of search queries in the first language [e.g. Japanese], and a request [e.g. translating search queries] for translating the first set more search queries in the first language to a second language [e.g. [0023]; Japanese, in the first language] different from the first language;
providing the one or more prompts for execution by the machine-learned language model [e.g. translate the search queries];
receiving a response generated by executing the machine-learned language model on the prompt, the response including a second set of search queries in the second language [e.g. FIG.6-7; translate the input queries to the queries in Japanese];
accessing a first set of features [e.g. FIG. 6; search query, language on search query, and language information on search results] generated based on user interactions [e.g. FIG. 1 and 6; a user terminal] with results of the first set of search queries in the first language [e.g. English or other language rather than Japanese];
generating a second set of features [e.g. FIG. 5-6; a translated or re-translated search query in the first language; for the second set of search queries based on the first set of features; the search condition converted to the first language as a feature value and the searched information in the first language]; and
training a machine-learning search model [e.g. model training]based on the second set of search queries and the second set of features [e.g. FIG. 2 and 6-7; searching information in the first language by inputting the search condition converted to the first language to the machine learning model; and generating training data with the search condition converted to the first language as a feature value and the searched information in the first language as a correct answer label].
It is noted that Oshio differs to the present invention in that Oshio fails to explicitly disclose context of a first set of search queries and deploying the trained machine-learning model online.
However, LAMBA teaches the well-known concept of obtaining a first of search for a first set of search queries [e.g. FIG. 2-3; text queries; e.g. [0027]; information day of the week associated with the text queries]] in a first language [e.g. non-English languages] and context of the first set of search queries [e.g. contextual information] for training a language model [e.g. FIG. 2-4; [0003-0006]; The trained machine learning model makes routing predictions for text queries for the language having sparse training text data].
It would have been obvious to a person of ordinary skill in the art before the effective filing date of the invention to modify the computing system disclosed by Oshio to exploit the well-known obtaining user queries for training a language model technique taught by LAMBA as above, in order to provide routing predictions for text queries for the language having sparse training text data [See LAMBA; [0003-0006]].
Moreover, Hwang teaches the well-known concept of an online system deploying the trained machine-learning search model [e.g. FIG. 1 and 6; trained model] onto an online platform servicing search queries [e.g. identify a task to perform a search] in the second language [e.g. FIG. 1-2 and 6; using a model trained from training data that is created from training data in another language; translated language].
It would have been obvious to a person of ordinary skill in the art before the effective filing date of the invention to modify the computing system disclosed by Oshio to exploit the well-known obtaining user queries for training a language model technique taught by LAMBA and the well-known deploying a trained model onto online platform technique taught by Hwang as above, in order to provide routing predictions for text queries for the language having sparse training text data [See LAMBA; [0003-0006]] and reduced time consuming for collecting data [See Hwang; [0001]].
Regarding claim 13, this is a non-transitory computer-readable storage medium that includes same limitation as in claim 1 above, the rejection of which are incorporated herein. Furthermore, Oshio, LAMBA and Hwang disclose a non-transitory computer-readable storage medium [e.g. Oshio: FIG. 1 and 8; [0005]], having instructions[e.g. program] encoded thereon that, when executed by one or more processors [e.g. [0044]; processors], cause the one or more processors to perform steps.
Regarding claim 20, this is a system that includes same limitation as in claim 1 above, the rejection of which are incorporated herein.
Claim(s) 2-12 and 14-19 is/are rejected under 35 U.S.C. 103 as being unpatentable over Oshio et al (US 20220335046 A1) in view of LAMBA et al (US 20230281399 A1), Hwang et al (US 20150127319 A1) and Wang et al (US 20230146336 A1).
Regarding claim 2, Oshio, LAMBA and Hwang further disclose accessing the first set of features [e.g. FIG.1; LAMBA: contextual features] comprises, for a search query in the first set [e.g. LAMBA: FIG. 1-2; text queries in a language], accessing one or more product identifiers (IDs) [e.g. LAMBA: [0020 and 0027]; product ID], but Oshio, LAMBA and Hwang fail to explicitly disclose interaction rates associated with each product ID.
However, Wang teaches the well-known concept of accessing a search query in the first set [e.g. FIG. 3-4; received query], accessing one or more product identifiers (IDs) for products [e.g. FIG. 4-5; item Identification or name] retrieved for the search query and interaction rates associated with each product ID [e.g. [0017-0020 and 0039]; shopper name, gender, rating, previous shopping history; the rates which the item and the additional item co-occur in orders previously received from users of the online concierge system].
It would have been obvious to a person of ordinary skill in the art before the effective filing date of the invention to modify the computing system disclosed by Oshio to exploit the well-known obtaining user queries for training a language model technique taught by LAMBA, the well-known deploying a trained model onto online platform technique taught by Hwang and the well-known concept of retrieving an item from a database technique taught by Wang as above, in order to provide routing predictions for text queries for the language having sparse training text data [See LAMBA; [0003-0006]], reduced time consuming for collecting data [See Hwang; [0001]] and simplified retrieval of items from a database [See Wang: abstract].
Regarding claim 3, Oshio, LAMBA, Hwang and Wang further disclose obtaining the context of the first set of search queries includes for a search query in the first set [e.g. Oshio: FIG. 5-6; LAMBA: FIG. 2-4], one or a combination of: a type of retailer store the search query was submitted for [e.g. Wang: warehouse], user profile information of a user who submitted the search query [e.g. LAMBA: FIG. 2 and 5; [0027]; user ID] .
It would have been obvious to a person of ordinary skill in the art before the effective filing date of the invention to modify the computing system disclosed by Oshio to exploit the well-known obtaining user queries for training a language model technique taught by LAMBA, the well-known deploying a trained model onto online platform technique taught by Hwang and the well-known concept of retrieving an item from a database technique taught by Wang as above, in order to provide routing predictions for text queries for the language having sparse training text data [See LAMBA; [0003-0006]], reduced time consuming for collecting data [See Hwang; [0001]] and simplified retrieval of items from a database [See Wang: abstract].
Regarding claim 4, Oshio, LAMBA, Hwang and Wang further disclose the machine-learned language model generates a plurality of translations for each search query in the first set of one or more search queries based on the context [e.g. Oshio: FIG. 5-6; LAMBA: FIG. 2-4].
It would have been obvious to a person of ordinary skill in the art before the effective filing date of the invention to modify the computing system disclosed by Oshio to exploit the well-known obtaining user queries for training a language model technique taught by LAMBA, the well-known deploying a trained model onto online platform technique taught by Hwang and the well-known concept of retrieving an item from a database technique taught by Wang as above, in order to provide routing predictions for text queries for the language having sparse training text data [See LAMBA; [0003-0006]], reduced time consuming for collecting data [See Hwang; [0001]] and simplified retrieval of items from a database [See Wang: abstract].
Regarding claim 5, Oshio, LAMBA, Hwang and Wang further disclose collecting the first search log for the first set of one or more search queries in the first language and context of the first set of one or more search queries from a first online platform in the first language [e.g. Oshio: FIG. 5-6; LAMBA: FIG. 1 and 3; [0020]; platform type; e.g. English user input] , and deploying the machine-learning search model in the second language onto a second online platform in a second language [e.g. Oshio: FIG. 5-6; LAMBA: 1 and 3; Non-English user input] wherein deploying the trained machine-learning search model comprises deploying the trained machine-learning search model onto a second online platform different from the first online platform [e.g. Oshio: FIG. 5-6; LAMBA: 1 and 3; Hwang: FIG. 5-7].
It would have been obvious to a person of ordinary skill in the art before the effective filing date of the invention to modify the computing system disclosed by Oshio to exploit the well-known obtaining user queries for training a language model technique taught by LAMBA, the well-known deploying a trained model onto online platform technique taught by Hwang and the well-known concept of retrieving an item from a database technique taught by Wang as above, in order to provide routing predictions for text queries for the language having sparse training text data [See LAMBA; [0003-0006]], reduced time consuming for collecting data [See Hwang; [0001]] and simplified retrieval of items from a database [See Wang: abstract].
Regarding claim 6, Oshio, LAMBA, Hwang and Wang further disclose collecting a third search log for a third set of one or more search queries in the second language and context of the third set of one or more search queries from the second online platform in the first language [e.g. Oshio: FIG. 5-6; LAMBA: FIG. 1 and 3; Non-English user input]; and fine-tuning the machine-learning search model based on the third search log and corresponding context [e.g. Oshio: FIG. 5-6; generate the training data and re-train the machine learning model; LAMBA: 1 and 3; generating a training text corpus in multiple languages].
It would have been obvious to a person of ordinary skill in the art before the effective filing date of the invention to modify the computing system disclosed by Oshio to exploit the well-known obtaining user queries for training a language model technique taught by LAMBA, the well-known deploying a trained model onto online platform technique taught by Hwang and the well-known concept of retrieving an item from a database technique taught by Wang as above, in order to provide routing predictions for text queries for the language having sparse training text data [See LAMBA; [0003-0006]], reduced time consuming for collecting data [See Hwang; [0001]] and simplified retrieval of items from a database [See Wang: abstract].
Regarding claim 7, Oshio, LAMBA, Hwang and Wang further disclose periodically fine-tuning the machine-learning search model based on a new search log and corresponding context generated by recent user interactions with the second online platform in the second language [e.g. Oshio: FIG. 5-6; generate the training data and re-train the machine learning model; LAMBA: FIG. 1 and 3; generating a training text corpus in multiple languages].
It would have been obvious to a person of ordinary skill in the art before the effective filing date of the invention to modify the computing system disclosed by Oshio to exploit the well-known obtaining user queries for training a language model technique taught by LAMBA, the well-known deploying a trained model onto online platform technique taught by Hwang and the well-known concept of retrieving an item from a database technique taught by Wang as above, in order to provide routing predictions for text queries for the language having sparse training text data [See LAMBA; [0003-0006]], reduced time consuming for collecting data [See Hwang; [0001]] and simplified retrieval of items from a database [See Wang: abstract].
Regarding claim 8, Oshio, LAMBA, Hwang and Wang further disclose the machine-learning search model includes a query understanding model [e.g. LAMBA: FIG. 2-4; deep learning model to handle different languages] configured to interpret a search query to determine context of the search query, the context indicating a category of products the search query is intended to search [e.g. Oshio: FIG. 5-6; LAMBA: FIG. 1 and 3; product SKU; Wang: FIG. 2 and 4-5; item attributes; [0004]; a category of the item].
It would have been obvious to a person of ordinary skill in the art before the effective filing date of the invention to modify the computing system disclosed by Oshio to exploit the well-known obtaining user queries for training a language model technique taught by LAMBA, the well-known deploying a trained model onto online platform technique taught by Hwang and the well-known concept of retrieving an item from a database technique taught by Wang as above, in order to provide routing predictions for text queries for the language having sparse training text data [See LAMBA; [0003-0006]], reduced time consuming for collecting data [See Hwang; [0001]] and simplified retrieval of items from a database [See Wang: abstract].
Regarding claim 9, Oshio, LAMBA, Hwang and Wang further disclose the machine-learning search model includes a recall model configured to retrieve a set of products from a catalog database based on a search query [e.g. Wang: FIG. 2 and 4-5; [0039]; retrieving shopping history].
It would have been obvious to a person of ordinary skill in the art before the effective filing date of the invention to modify the computing system disclosed by Oshio to exploit the well-known obtaining user queries for training a language model technique taught by LAMBA, the well-known deploying a trained model onto online platform technique taught by Hwang and the well-known concept of retrieving an item from a database technique taught by Wang as above, in order to provide routing predictions for text queries for the language having sparse training text data [See LAMBA; [0003-0006]], reduced time consuming for collecting data [See Hwang; [0001]] and simplified retrieval of items from a database [See Wang: abstract].
Regarding claim 10, Oshio, LAMBA, Hwang and Wang further disclose the machine-learning search model includes a click through rate model configured to determine a probability that a user will click on a particular product when the particular product is shown in a search result [e.g. Oshio: FIG. 5-6; Wang: FIG. 5-6; the affinity score between the item corresponding to the item identifier and the query term may be determined from rates at which users selected the item via the online concierge system after providing the query term to the online concierge system].
It would have been obvious to a person of ordinary skill in the art before the effective filing date of the invention to modify the computing system disclosed by Oshio to exploit the well-known obtaining user queries for training a language model technique taught by LAMBA, the well-known deploying a trained model onto online platform technique taught by Hwang and the well-known concept of retrieving an item from a database technique taught by Wang as above, in order to provide routing predictions for text queries for the language having sparse training text data [See LAMBA; [0003-0006]], reduced time consuming for collecting data [See Hwang; [0001]] and simplified retrieval of items from a database [See Wang: abstract].
Regarding claim 11, Oshio, LAMBA, Hwang and Wang further disclose the machine-learning search model includes a ranking model configured to rank search results based on one or more criteria, the one or more criteria including click through rates of products in the search result [e.g. Oshio: FIG. 5-6; Wang: FIG. 4 and 7; he model 700 ranks the items based on the predicted similarity; the online concierge system ranks items based on the predicted similarity of their corresponding item identifier to the embedding for the term of the query].
It would have been obvious to a person of ordinary skill in the art before the effective filing date of the invention to modify the computing system disclosed by Oshio to exploit the well-known obtaining user queries for training a language model technique taught by LAMBA, the well-known deploying a trained model onto online platform technique taught by Hwang and the well-known concept of retrieving an item from a database technique taught by Wang as above, in order to provide routing predictions for text queries for the language having sparse training text data [See LAMBA; [0003-0006]], reduced time consuming for collecting data [See Hwang; [0001]] and simplified retrieval of items from a database [See Wang: abstract].
Regarding claim 12, Oshio, LAMBA, Hwang and Wang further disclose applying the trained machine-learning search model to process new search queries in the second language [e.g. Oshio: FIG. 5-6; Wang: FIG. 5-6: LAMBA: 1 and 3; Non-English or English user input], wherein applying the trained machine-learning search model produces search results based on the second set of features [e.g. Oshio: FIG. 5-6; LAMBA: 2-3]; collecting user feedback on the search results provided by the trained machine-learning search model [e.g. Oshio: FIG. 5-6; feedback data and training data; Wang: FIG. 2 and 4-6], wherein the user feedback comprises one or more interaction rates associated with the search results [e.g. Wang: [0017-0020 and 0039]; shopper name, gender, rating, previous shopping history; the rates which the item and the additional item co-occur in orders previously received from users of the online concierge system]; and refining the trained machine-learning search model based on the user feedback to adjust model parameters and improve subsequent search result accuracy and relevance [e.g. Oshio: FIG. 5-6; Wang: FIG. 5-6; [0013-0014]; modifying parameters to reduce one or more items; the model is refined using multi-task training where a task is trained to predict scores for items from the received query].
It would have been obvious to a person of ordinary skill in the art before the effective filing date of the invention to modify the computing system disclosed by Oshio to exploit the well-known obtaining user queries for training a language model technique taught by LAMBA, the well-known deploying a trained model onto online platform technique taught by Hwang and the well-known concept of retrieving an item from a database technique taught by Wang as above, in order to provide routing predictions for text queries for the language having sparse training text data [See LAMBA; [0003-0006]], reduced time consuming for collecting data [See Hwang; [0001]] and simplified retrieval of items from a database [See Wang: abstract].
Regarding claim 14-18, this is a non-transitory computer-readable storage medium that includes same limitation as in claim 2-6 above, the rejection of which are incorporated herein.
Regarding claim 19, this is a non-transitory computer-readable storage medium that includes same limitation as in claim 5 and 12 together above, the rejection of which are incorporated herein.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
GANDHI et al (US 20240378399 A1).
Wu et al (US 20230267285 A1).
THIS ACTION IS MADE FINAL. 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 nonprovisional extension fee (37 CFR 1.17(a)) 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 ZHUBING REN whose telephone number is (571)272-2788. The examiner can normally be reached Monday-Friday 9am-5pm.
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, Richemond Dorvil can be reached at 571-272-7602. 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.
/ZHUBING REN/Primary Examiner, Art Unit 2658
Prosecution Insights