ETAILED 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 .
Response to Arguments
Applicant's arguments filed on have been fully considered but they are not persuasive.
101 Rejections:
With respect to Applicant’s argument that “According to the Office, "determining... a second set of candidate languages based on one or more information resources... recites a mental process..." (Instant Action, p. 10). The Office further alleges, "Viewing the additional limitations together and the claim as a whole, nothing provides integration into a practical application," (Id. at p. 12). As amended, claim 1 recites, at least in part, "identifying... using the language recognition model, languages associated with one or more information resources associated with the one or more keywords; determining, by the data processing system, a second set of candidate languages based on the identified languages associated with the one or more information resources." As such, Applicant respectfully submits that amended claim 1 is patent eligible under 35 U.S.C. § 101”, Examiner respectfully disagrees.
Examiner cites that the amended limitations “identifying, process because human mind can identify languages associated with information resources associated with one or more keywords by evaluation and judgement of data.
-“determining,
The limitation “by the data processing system”, which is a high-level recitation of a generic computer components and represent mere instructions to apply the judicial exception on a computer as in MPEP 2106.05(f), which does not provide integration into a practical application.
The limitation “using a language recognition model” is generally linking the use of the judicial exception to a particular technological environment or field of use by limiting it to a particular data source or type. See MPEP § 2106.05(h) (technical environment of ML based language data model) and/or apply it limitation of using the computer as a tool, i.e ML-based language data model, to perform the judicial exception. See MPEP (2106.05(f)).
Therefore, the claims are not patent eligible.
With respect to Applicant’s argument that “In particular, Applicant respectfully contends that the instant claims include additional elements that amount to significantly more than an alleged abstract idea at least by improving the functioning of language recognition techniques. According to the instant specification, "By using multiple factors... the accuracy of the languages predicted to be used by the user can be significantly increased, as much as 70-90% in comparison to using only the language declared by the user or derived from the keywords of the query," (Instant Specification, para. [0029]). Further, "the set of content items from which to select and provide can be expanded to include multiple languages that are determined with greater accuracy and precision... [which] may lead to a decreased consumption of computing and network resources, with the user making less queries to find relevant content," (Id.). Moreover, amended claim 1 provides such an improvement at least by identifying, using a language recognition model, languages associated with information resource(s) and determining a second set of candidate languages based on the identified languages. Applicant respectfully contends that the instant claims therefore cause the improvements to the functionality as described above”, Examiner respectfully disagrees.
Applicant argues that amended claim 1 improves the functioning of language-recognition techniques because using multiple factors, including languages identified from information resources, increases the accuracy of the languages predicted to be used by the user. Examiner respectfully disagrees. The claimed limitations do not recite an improvement to technology. Rather, the claim uses the language recognition model to identify languages associated with information resources and uses the identified languages as additional information for determining a second set of candidate languages. As discussed above, the recitation of the language recognition model merely limits the claimed analysis to a particular technological environment and uses the model as a tool to perform mental processes identified above. Thus, the additional elements do not, merely specifying the model and information resources used in performing the analysis, integrate the judicial exception into a practical application. See MPEP §§ 2106.05(f) and 2106.05(h).
Accordingly, the asserted increase in prediction accuracy does not demonstrate an improvement to the technology.
With respect to Applicant’s argument that “Further, According to the Office, "'[U]sing a language recognition model' ... is generally linking the use of the judicial exception to a particular technological environment or field of use by limiting it to a particular data source or type," (Instant Action, p. 12). Applicant respectfully disagrees. In particular, as discussed during the Interview, Applicant respectfully contends that the corresponding elements of amended claim 1 are not merely linking the use of a judicial exception or well-understood, routine, or conventional (WURC) activity. In particular,9
…. therefore, to identify languages of specific information resources associated with particular keywords (e.g., of a search) and subsequently adding the languages to user language preferences is a technical solution to the particular technical problem arising in the technical environment. The particular steps of claim 1 therefore comprise unconventional steps "that confine the claim to a particular useful application..." as recited in the MPEP (MPEP § 2106.05(I)(A)(v))”, Examiner respectfully disagrees.
Applicant argues that the amended limitations do not merely link the judicial exception to a particular technological environment or field of use, but instead provide a technical solution to a technical problem arising from Internet search results, which Applicant characterizes as uncategorized and rapidly updating. Applicant further argues that the recited limitations constitute non-conventional and non-generic steps that confine the claim to a particular useful application. Examiner respectfully disagrees. The claim does not recite an improvement to the manner in which Internet information is categorized, updated, searched, or retrieved, nor does the claim recite an improvement to the operation or functionality of the language recognition model itself. Rather, the language recognition model is used as a tool to identify languages associated with information resources associated with one or more keywords, and the identified languages are subsequently used in determining a second set of candidate languages. Thus, the additional elements use the recited computer-based tools and information resources to perform the claimed analysis rather than improving the operation of the language recognition model, Internet search technology, or the computer itself. Accordingly, considering the additional elements individually and in combination with the judicial exception, the additional elements do not integrate the judicial exception into a practical application. See MPEP §§ 2106.05(f) and 2106.05(h). Further, to the extent Applicant’s argument that the recited elements are non-conventional is directed to Step 2B, as discussed in the Office Action, the additional computer elements perform well-understood, routine, and conventional functions of receiving, and processing information and do not, individually or as an ordered combination, amount to significantly more than the judicial exception. See MPEP § 2106.05(d)(II).
Applicant further argues that independent claim 11 and the dependent claims are patent eligible for reasons similar to those discussed with respect to claim 1. Examiner respectfully disagrees. For the same reasons discussed above with respect to claim 1, the corresponding limitations of independent claim 11 do not integrate the judicial exception into a practical application and, when considered individually and as an ordered combination, the additional elements do not amount to significantly more than the judicial exception. Accordingly, Applicant’s arguments with respect to independent claim 11 and the dependent claims are not persuasive.
103 Rejection:
Applicant’s arguments, with respect to the rejection(s) of claim(s) 1-20 under 103 have been fully considered and are persuasive. However, upon further consideration, a new ground(s) of rejection is made in view of Datta (US 2007/0288450).
Double Patenting
The nonstatutory double patenting rejection is based on a judicially created doctrine grounded in public policy (a policy reflected in the statute) so as to prevent the unjustified or improper timewise extension of the “right to exclude” granted by a patent and to prevent possible harassment by multiple assignees. A nonstatutory double patenting rejection is appropriate where the conflicting claims are not identical, but at least one examined application claim is not patentably distinct from the reference claim(s) because the examined application claim is either anticipated by, or would have been obvious over, the reference claim(s). See, e.g., In re Berg, 140 F.3d 1428, 46 USPQ2d 1226 (Fed. Cir. 1998); In re Goodman, 11 F.3d 1046, 29 USPQ2d 2010 (Fed. Cir. 1993); In re Longi, 759 F.2d 887, 225 USPQ 645 (Fed. Cir. 1985); In re Van Ornum, 686 F.2d 937, 214 USPQ 761 (CCPA 1982); In re Vogel, 422 F.2d 438, 164 USPQ 619 (CCPA 1970); In re Thorington, 418 F.2d 528, 163 USPQ 644 (CCPA 1969).
A timely filed terminal disclaimer in compliance with 37 CFR 1.321(c) or 1.321(d) may be used to overcome an actual or provisional rejection based on nonstatutory double patenting provided the reference application or patent either is shown to be commonly owned with the examined application, or claims an invention made as a result of activities undertaken within the scope of a joint research agreement. See MPEP § 717.02 for applications subject to examination under the first inventor to file provisions of the AIA as explained in MPEP § 2159. See MPEP § 2146 et seq. for applications not subject to examination under the first inventor to file provisions of the AIA . A terminal disclaimer must be signed in compliance with 37 CFR 1.321(b).
The filing of a terminal disclaimer by itself is not a complete reply to a nonstatutory double patenting (NSDP) rejection. A complete reply requires that the terminal disclaimer be accompanied by a reply requesting reconsideration of the prior Office action. Even where the NSDP rejection is provisional the reply must be complete. See MPEP § 804, subsection I.B.1. For a reply to a non-final Office action, see 37 CFR 1.111(a). For a reply to final Office action, see 37 CFR 1.113(c). A request for reconsideration while not provided for in 37 CFR 1.113(c) may be filed after final for consideration. See MPEP §§ 706.07(e) and 714.13.
The USPTO Internet website contains terminal disclaimer forms which may be used. Please visit www.uspto.gov/patent/patents-forms. The actual filing date of the application in which the form is filed determines what form (e.g., PTO/SB/25, PTO/SB/26, PTO/AIA /25, or PTO/AIA /26) should be used. A web-based eTerminal Disclaimer may be filled out completely online using web-screens. An eTerminal Disclaimer that meets all requirements is auto-processed and approved immediately upon submission. For more information about eTerminal Disclaimers, refer to www.uspto.gov/patents/apply/applying-online/eterminal-disclaimer.
Claims 1-20 are rejected on the ground of nonstatutory double patenting as being unpatentable over claims 1-22 of U.S. Patent No. 12, 182, 213 (hereinafter ‘213) in view of Sonntag et al. (US 2021/0334299) and in view of Datta et al. (US 2007/0288450).
Instant Application
US Patent No. 12, 182, 213
Claim 1: A method, comprising: receiving, by a data processing system having one or more processors, from a client device, a request for content identifying an account profile and including one or more keywords;
Claim 1: A method, comprising: receiving, by a data processing system having one or more processors, from a client device, a request for content identifying an account profile;
Claim 7: (iii) one or more keywords included in the request for content;
determining, by the data processing system using a log record identifying a browsing history of the account profile, a first set of candidate languages from a plurality of languages by analyzing the log record using a language recognition model, wherein the language recognition model is trained according to a training dataset including corpuses of text for each language of the plurality of languages;
determining, by the data processing system using a log record identifying a browsing history of the account profile, a first set of candidate languages from a plurality of languages by analyzing the log record using a language recognition model, wherein the language recognition model is trained according to a training dataset including corpuses of text for each language of the plurality of languages by:
applying each of the corpuses of text for each language of the plurality of languages to the training dataset to generate a set of results corresponding to result languages of the plurality of languages, generating a result error by comparing each of the result languages to a labeled language for each of the corpuses,
and modifying one or more weights of the language recognition model based on the result error;
determining, by the data processing system, a second set of candidate languages based on one or more information resources associated with the one or more keywords;
determining, by the data processing system, a second set of candidate languages from the plurality of languages based on a language setting of the account profile;
identifying, by the data processing system, a set of languages included in both the first set of candidate languages and the second set of candidate languages, the set of languages including a first language and a second language from the plurality of the languages;
identifying, by the data processing system and using the language recognition model, languages associated with one or more information resources associated with the one or more keywords;
determining, by the data processing system, a second set of candidate languages based on the identified languages associated with the one or more information resources associated with the one or more keywords;
and storing, by the data processing system in one or more data structures, an association among the account profile, the first language, and the second language.
calculating, by the data processing system, confidence scores for at least some of the second set of candidate languages; and
updating, by the data processing system, the first set of candidate languages based on the confidence scores for the at least some of the second set of candidate languages.
With respect to claim 1, US Patent '213 teaches all the limitations of claim 1,
except for the limitation "calculating, by the data processing system, confidence scores for at least some of the second set of candidate languages; updating, by the data processing system, the first set of candidate languages based on the confidence scores for the at least some of the second set of candidate languages and identifying, by the data processing system and using the language recognition model, languages associated with one or more information resources associated with the one or more keywords; determining, by the data processing system, a second set of candidate languages based on the identified languages associated with the one or more information resources associated with the one or more keywords”.
However, Sonntag teaches calculating, by the data processing system, confidence scores for at least some of the second set of candidate languages ([0007, A weighted score is determined by at least the respective language confidence score and the respective match score. The weighted score is used as a basis by the operations to determine at least one query language for the user query.], [0025, candidate languages associated with confidence scores that meet a confidence threshold may be used to determine which language-specific databases to search], calculating confidence scores for plurality of languages (includes second set of candidate languages);
updating, by the data processing system, the first set of candidate languages based on the confidence scores for the at least some of the second set of candidate languages ([0004, A weighted score is determined by at least the respective language confidence score and the respective match score], [0063, demoted score may remove a particular language even if the confidence score without the demote value meets a threshold or ranking criteria. A language-specific database of a language with a boosted score that meets the searching criteria may be searched], [0073, a confidence score for a particular search query, for example, may be 90% for French and 60% for Spanish based on the machine learning model output. A top three matches in a French database have match scores of 500 (F1), 450 (F2), and 200 (F3) respectively, and a top three matches in a Spanish database have match scores of 750 (S1), 400 (S2) and 250 (S3), respectively… F1 and S1, F2, and S2], the confidence scores for multiple languages are used to compute weighted scores, which are then used to determine ranking results, the weighted scores are boosted or demoted, resulting in changes in ranking results, i.e. French before Spanish. Modifying the weighted scores changes ranking between languages and because selection is based on these scores, modifying them updates the selected languages based on their associated confidence scores).
Therefore, it would have been obvious to one of the ordinary skills in the art before the elective filing date to incorporate functionalities of Sonntag into the invention of US Patent ‘213 to improve language selection by enabling more accurate identification of the most suitable language (Sonntag, [0021, The online game system employs a language analysis system to detect potential languages of a user query and to assist in efficient and reliable searching for results to formulate the response]).
However, Datta teaches identifying, by the data processing system and using the language recognition model, languages associated with one or more information resources associated with the one or more keywords ([0024, If "elephant" is a common-form entry in the synonyms map, then the variants "elephant", "elephant" and "elephant" would be associated with that entry as values, if these variants are found in the corpus… Each value also includes additional information: the languages of the documents in which instances of the variant appear, and the number of times the variant appears in that language. Continuing the example, "elephant" might be found in the corpus 90 times among documents recognized as being English and 300 times among documents recognized as being French], multiple languages (French, English) associated with the documents (information resources) are determined by the using the Naïve Bayes models (language recognition model), wherein the documents are associated with keywords, for example, determining how many times the keyword elephant appear in the document (resources));
determining, by the data processing system, a second set of candidate languages based on the identified languages associated with the one or more information resources ([0043, The process 200 also removes any languages associated with a variant that have a frequency that does not exceed a pre-defined absolute threshold (step 230)], [0047, From each variant in the synonyms map, the process 200 removes any languages where the language's relative frequency does not meet a pre-defined relative threshold (e.g., 10%) (step 260). The same threshold applies to all variants and all languages], multiple languages associated with the documents (information resources) are determined, and languages that do not satisfy predetermined frequency thresholds are removed. The languages remaining after the removal/filtering are the “second set of candidate languages”).
Therefore, it would have been obvious to one of the ordinary skills in the art before the elective filing date to incorporate functionalities of Datta into the invention of US Patent ‘213/Sontag to allow the system to prioritize languages that are more likely to be relevant to the user, thereby improving search results and overall system performance (Datta, [0056, adjustment of the query classifier can be iteratively repeated to further improve the accuracy.]).
Claim 2 corresponds to claim 2 of US Patent ‘213.
Claim 3 corresponds to claim 3 of US Patent ‘213.
Claim 4 corresponds to claim 3 of US Patent ‘213.
Claim 5 corresponds to claim 4 of US Patent ‘213.
Claim 6 corresponds to claim 5 of US Patent ‘213.
With respect to claim 7, US Patent '213 teaches all the limitations of claim 7,
except for the limitation "updating, by the data processing system, the first set of candidate languages based on the third set of candidate languages”.
However, Sonntag teaches updating, by the data processing system, the first set of candidate languages based on the third set of candidate languages ([0061, Language-specific databases for the top designated number of ranked languages may be searched. In some implementations, language-specific databases of candidate languages that meet a threshold confidence score are used in the search.], [0063, resulting boosted score of a particular language may reach a confidence threshold or ranking, even if the confidence score without the boost value fails to meet a threshold or ranking criteria], languages are included/removed based on the confidence scores of other languages (updating based on other languages)).
Therefore, it would have been obvious to one of the ordinary skills in the art before the elective filing date to incorporate functionalities of Sonntag into the invention of US Patent ‘213 to improve language selection by enabling more accurate identification of the most suitable language (Sonntag, [0021, The online game system employs a language analysis system to detect potential languages of a user query and to assist in efficient and reliable searching for results to formulate the response]).
Claim 8 corresponds to claim 8 of US Patent ‘213.
Claim 9 corresponds to claim 22 of US Patent ‘213.
Claim 10 corresponds to claim 10 of US Patent ‘213.
Claim 11 of the instant application corresponds to claim 11 of US Patent ‘213 with differences being similar to differences in claim 1 discussed above, the differences are obvious for the same rational discussed above.
Claim 12 corresponds to claim 12 of US Patent ‘213.
Claim 13 corresponds to claim 13 of US Patent ‘213.
Claim 14 corresponds to claim 13 of US Patent ‘213.
Claim 15 corresponds to claim 14 of US Patent ‘213.
Claim 16 corresponds to claim 15 of US Patent ‘213.
Claim 17 of the instant application corresponds to claim 17 of US Patent ‘213 with differences being similar to differences in claim 7 discussed above, the differences are obvious for the same rational discussed above.
Claim 18 corresponds to claim 18 of US Patent ‘213.
Claim 19 corresponds to claim 22 of US Patent ‘213.
Claim 20 corresponds to claim 21 of US Patent ‘213.
Claim Rejections - 35 USC § 101
35 U.S.C. 101 reads as follows:
Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.
Claims 1-20 are rejected under 35 U.S.C. 101 because of the following reasons:
Claims 1, 11:
At Step 1:
The claims are directed to a “method” and "system" and thus directed to a statutory category.
At Step 2A, Prong One:
The claim recites the following limitations directed to an abstract idea:
-“determining, of languages by analyzing the log record by identifying users browsing histories associated with a user account by analyzing log record by evaluation and judgement of data.
-“identifying,
-“determining,
-“calculating,
-“updating,
At Step 2A, Prong Two:
The claim recites the following additional elements:
-“by the data processing system” which is a high-level recitation of a generic computer components and represent mere instructions to apply the judicial exception on a computer as in MPEP 2106.05(f), which does not provide integration into a practical application.
-“receiving, by a data processing system having one or more processors, from a client device, a request for content identifying an account profile and including one or more keywords” is insignificant extra-solution activity as mere data gathering such as 'obtaining information'. See MPEP 2106.05(g).
-“using a language recognition model”, “wherein the language recognition model is trained according to a training dataset including corpuses of text for each language of the plurality of languages” is generally linking the use of the judicial exception to a particular technological environment or field of use by limiting it to a particular data source or type. See MPEP § 2106.05(h) (technical environment of ML based language data model) and/or apply it limitation of using the computer as a tool, i.e ML-based language data model, to perform the judicial exception. See MPEP (2106.05(f)).
Viewing the additional limitations together and the claim as a whole, nothing provides integration into a practical application.
At Step 2B:
The conclusions for the mere implementation using a computer are carried over and does not provide significantly more.
-“receiving, by a data processing system having one or more processors, from a client device, a request for content identifying an account profile and including one or more keywords” is well-understood, routine and conventional activities (WURC) as evidenced by the court cases cited in MPEP 2106.05(d)(II) by at least "i. Receiving or transmitting data over a network, e.g., using the Internet to gather data, Symantec, 838 F.3d at 1321, … buySAFE, Inc. v. Google, Inc., 765 F.3d 1350, 1355, 112 USPQ2d 1093, 1096 (Fed. Cir. 2014) (computer receives and sends information over a network)".
-“using a language recognition model”, “wherein the language recognition model is trained according to a training dataset including corpuses of text for each language of the plurality of languages” is generally linking the use of the judicial exception to a particular technological environment or field of use by limiting it to a particular data source or type. See MPEP § 2106.05(h) (technical environment of ML based language data model) and/or apply it limitation of using the computer as a tool, i.e ML-based language data model, to perform the judicial exception. See MPEP (2106.05(f)).
Accordingly, at step 2B, these additional elements, both individually and in combination, do not amount to significantly more than the judicial exception. See MPEP § 2106.05. Therefore, the claim is not eligible subject matter under 35 U.S.C. 101.
Claims 2, 12
At Step 2A, Prong One:
The claim recites the following limitations directed to an abstract idea:
-“generating, based on the occurrences of the first languages in the browsing histories by evaluation and judgment of data.
At Step 2A, Prong Two:
The claim recites the following additional elements:
-“wherein the confidence scores are second confidence scores” is generally linking the use of the judicial exception to a particular technological environment or field of use by limiting it to a particular data source or type. See MPEP § 2106.05(h) (technical environment of ML based language data model) and/or apply it limitation of using the computer as a tool, i.e. ML-based language data model, to perform the judicial exception. See MPEP (2106.05(f)).
Viewing the additional limitations together and the claim as a whole, nothing provides integration into a practical application.
At Step 2B:
The conclusions for the mere implementation using a computer are carried over and does not provide significantly more.
-“wherein the confidence scores are second confidence scores” is generally linking the use of the judicial exception to a particular technological environment or field of use by limiting it to a particular data source or type. See MPEP § 2106.05(h) (technical environment of ML based language data model) and/or apply it limitation of using the computer as a tool, i.e. ML-based language data model, to perform the judicial exception. See MPEP (2106.05(f)).
Accordingly, at step 2B, these additional elements, both individually and in combination, do not amount to significantly more than the judicial exception. See MPEP § 2106.05. Therefore, the claim is not eligible subject matter under 35 U.S.C. 101.
Claims 3, 13:
At Step 2A, Prong One:
The claim recites the following limitations directed to an abstract idea:
-“including,
Claims 4, 14:
At Step 2A, Prong One:
The claim recites the following limitations directed to an abstract idea:
-“including,
Claims 5, 15:
At Step 2A, Prong One:
The claim recites the following limitations directed to an abstract idea:
-“identifying,
At Step 2A, Prong Two:
The claim recites the following additional elements:
-“providing, by the data processing system to the client device, a content item selected from one of the first plurality of content items and the second plurality of content items, the content item in one of the first language or the second language”
is- insignificant extra-solution activity as mere data outputting. See MPEP 2106.05(g).
Viewing the additional limitations together and the claim as a whole, nothing provides integration into a practical application.
At Step 2B:
The conclusions for the mere implementation using a computer are carried over and does not provide significantly more.
-“providing, by the data processing system to the client device, a content item selected from one of the first plurality of content items and the second plurality of content items, the content item in one of the first language or the second language” is well-understood, routine and conventional activities (WURC), See MPEP 2106.05(d)(II) by at least "iv. Presenting offers and gathering statistics, OIP Techs., 788 F.3d at 1362-63, 115 USPQ2d at 1092-9".
Accordingly, at step 2B, these additional elements, both individually and in combination, do not amount to significantly more than the judicial exception. See MPEP § 2106.05. Therefore, the claim is not eligible subject matter under 35 U.S.C. 101.
Claims 6, 16
At Step 2A, Prong One:
The claim recites the following limitations directed to an abstract idea:
-“identifying,
-“selecting,
At Step 2A, Prong Two:
The claim recites the following additional elements:
-“to provide to the client device in accordance to a content selection protocol, the content item in one of the first language or the second language” is insignificant extra-solution activity as mere data gathering such as 'outputting data'. See MPEP 2106.05(g).
Viewing the additional limitations together and the claim as a whole, nothing provides integration into a practical application.
At Step 2B:
The conclusions for the mere implementation using a computer are carried over and does not provide significantly more.
-“to provide to the client device in accordance to a content selection protocol, the content item in one of the first language or the second language” is well-understood, routine and conventional activities (WURC), See MPEP 2106.05(d)(II) by at least "iv. Presenting offers and gathering statistics, OIP Techs., 788 F.3d at 1362-63, 115 USPQ2d at 1092-9"
Accordingly, at step 2B, these additional elements, both individually and in combination, do not amount to significantly more than the judicial exception. See MPEP § 2106.05. Therefore, the claim is not eligible subject matter under 35 U.S.C. 101.
Claims 7, 17
At Step 2A, Prong One:
The claim recites the following limitations directed to an abstract idea:
-“identifying, from plurality of information resources/a language configuration/language setting by looking at the data by evaluation/observation and judgment of data.
-“and updating,
Claims 8, 18
At Step 2A, Prong Two:
The claim recites the following additional elements:
-“wherein the browsing history includes at least one of: a search query received from the client device, accessing of an information resource by the client device, and interaction with an element on information resource” is generally linking the use of the judicial exception to a particular technological environment or field of use by limiting it to a particular data source or type See MPEP § 2106.05(h) (technical environment of ML based language data model) and/or apply it limitation of using the computer as a tool, i.e ML-based language data model, to perform the judicial exception. See MPEP (2106.05(f)).
At Step 2B:
The conclusions for the mere implementation using a computer are carried over and does not provide significantly more.
-“wherein the browsing history includes at least one of: a search query received from the client device, accessing of an information resource by the client device, and interaction with an element on information resource” is generally linking the use of the judicial exception to a particular technological environment or field of use by limiting it to a particular data source or type See MPEP § 2106.05(h) (technical environment of ML based language data model) and/or apply it limitation of using the computer as a tool, i.e. ML-based language data model, to perform the judicial exception. See MPEP (2106.05(f)).
Accordingly, at step 2B, these additional elements, both individually and in combination, do not amount to significantly more than the judicial exception. See MPEP § 2106.05. Therefore, the claim is not eligible subject matter under 35 U.S.C. 101.
Claims 9, 19
At Step 2A, Prong Two:
The claim recites the following additional elements:
-“wherein the language recognition model is at least one of: (i) an artificial neural network, (ii) an n-gram model, (iii) a Bayesian network, (iv) a random forest model, (v) a support vector machine, or (vi) a decision tree model” is generally linking the use of the judicial exception to a particular technological environment or field of use by limiting it to a particular data source or type See MPEP § 2106.05(h) (technical environment of ML based language data model) and/or apply it limitation of using the computer as a tool, i.e. ML-based language data model, to perform the judicial exception. See MPEP (2106.05(f)).
At Step 2B:
The conclusions for the mere implementation using a computer are carried over and does not provide significantly more.
-“wherein the language recognition model is at least one of: (i) an artificial neural network, (ii) an n-gram model, (iii) a Bayesian network, (iv) a random forest model, (v) a support vector machine, or (vi) a decision tree model” is generally linking the use of the judicial exception to a particular technological environment or field of use by limiting it to a particular data source or type. See MPEP § 2106.05(h) (technical environment of ML based language data model) and/or apply it limitation of using the computer as a tool, i.e ML-based language data model, to perform the judicial exception. See MPEP (2106.05(f)).
Accordingly, at step 2B, these additional elements, both individually and in combination, do not amount to significantly more than the judicial exception. See MPEP § 2106.05. Therefore, the claim is not eligible subject matter under 35 U.S.C. 101.
Claims 10, 20
At Step 2A, Prong One:
The claim recites the following limitations directed to an abstract idea:
-“generating,
-“and modifying,
-“applying,
Claim Rejections - 35 USC § 103
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, 2, 3, 4, 5, 6, 7, 8, 9, 11, 12, 13, 14, 15, 16, 17, 18, 19 is/are rejected under 35 U.S.C. 103 as being unpatentable over Datar et al. (US 8,375,025) and in view of and in view of Sonntag et al. (US 2021/0334299) and in view of Datta et al. (US 2007/0288450).
With respect to claim 1, Datar teaches a method, comprising:
receiving, by a data processing system having one or more processors, from a client device, a request for content identifying an account profile and including one or more keywords ([col. 5, lines 1-5, “A search session can also be defined by a user indicating a beginning and an end of a search session (e.g., by logging into the search engine and logging out of the search engine)”], [col. 7, “A personal greeting 504 indicates that user John Doe (having an example username johndoe123) logged into the search engine so that the search engine can use any previously saved preferences. By selecting the sign out link 506, the user can log out of the search engine”], user login is received and causes the system to identify the user account/profile, the user also input search term (keyword) to search for content);
determining, by the data processing system using a log record identifying a browsing history of the account profile, a first candidate languages from a plurality of languages by analyzing the log record
determining, by the data processing system, a second candidate languages based on one or more information resources associated with the one or more keywords ([fig. 4, col. 6, lines 22-25, “FIG. 4 illustrates a set of example records that can be used in providing search results based on language selection statistics. The first example records 402 illustrates a recording scheme where the language preference 404 is stored in a selection record with a content item identifier, such as the network location (in this case a URL) 406.”], [fig. 6, “The other-language results section 604 can be provided to present results in other languages. A portion of the results page is shown as being reserved for one or more results for which the language selection statistics show a high degree of interest by users with the same language preference as the querying user”], the second preference language is identified from the users preference associated with the keywords “software download”, associated with the search results and the results are displayed in multiple languages);
calculating, by the data processing system, scores for at least some of the second candidate languages ([col. 8, lines 27-35, “the other-language results section 604 can exclude results categorized (by a crawler application, for example) as having the same language as the user's preferred language, and include one or more highest ranked results in another language based on ranking scores calculated using language selection statistics”], calculating scores for multiple languages (includes second language)); and
Datar does not explicitly teach a first set of candidate languages; using a language recognition model; wherein the language recognition model is trained according to a training dataset including corpuses of text for each language of the plurality of languages; a second set of candidate languages; calculating, by the data processing system, confidence scores for at least some of the second set of candidate languages; updating, by the data processing system, the first set of candidate languages based on the confidence scores for the at least some of the second set of candidate languages; identifying, by the data processing system and using the language recognition model, languages associated with one or more information resources associated with the one or more keywords; determining, by the data processing system, a second set of candidate languages based on the identified languages associated with the one or more information resources.
However, Sonntag teaches a first set of candidate languages and a second set of candidate languages ([0007, The at least one candidate language is used by the operations to select at least one language-specific databases to search. A response match is identified in the at least one language-specific databases and a respective match score is determined for each of the response match. A weighted score is determined by at least the respective language confidence score and the respective match score. The weighted score is used as a basis by the operations to determine at least one query language for the user query.], [0012, confidence scores for multiple languages is generated and multiple-language specific databases of the at least one language-specific game database corresponding to the respective confidence scores are searched based on the user query], at least one language is selected from the set of languages to search in the databases (first set of candidate languages); generating confidence scores for multiple languages by evaluating a plurality of possible languages (second set of candidate languages) );
using a language recognition model ([0030, The pre-processing enables the machine learning model 136 to recognize languages from a small number of input words in the user quay], the machine learning model is a language recognition model because it recognizes languages); wherein the language recognition model is trained according to a training dataset including corpuses of text for each language of the plurality of languages ([0004, The machine learning model is trained with at least one multilingual text corpus and game-related data. A respective language confidence score for each of the at least one candidate language is determined by applying the machine learning model], [0050, In block 302, the processed training data may be fed into an untrained machine learning model without the associated language labels for supervised training], [0052, In block 306, the current predicted language labels are compared with language labels that are associated with the training data prior to the training, such as the language labels from the pre-processing of the training data.], [0076, A listing of game information for a particular language may be validated by reference to supplemental language information], the language model is trained according to a training datasets which includes corpuses from multiple languages with a particular language label (for each language));
calculating, by the data processing system, confidence scores for at least some of the second set of candidate languages ([0007, A weighted score is determined by at least the respective language confidence score and the respective match score. The weighted score is used as a basis by the operations to determine at least one query language for the user query.], [0025, candidate languages associated with confidence scores that meet a confidence threshold may be used to determine which language-specific databases to search], calculating confidence scores for plurality of languages (includes second set of candidate languages));
updating, by the data processing system, the first set of candidate languages based on the confidence scores for the at least some of the second set of candidate languages ([0004, A weighted score is determined by at least the respective language confidence score and the respective match score], [0063, A resulting boosted score of a particular language may reach a confidence threshold or ranking, even if the confidence score without the boost value fails to meet a threshold or ranking criteria. A demoted score may remove a particular language even if the confidence score without the demote value meets a threshold or ranking criteria. A language-specific database of a language with a boosted score that meets the searching criteria may be searched], [0073, a confidence score for a particular search query, for example, may be 90% for French and 60% for Spanish based on the machine learning model output. A top three matches in a French database have match scores of 500 (F1), 450 (F2), and 200 (F3) respectively, and a top three matches in a Spanish database have match scores of 750 (S1), 400 (S2) and 250 (S3), respectively… F1 and S1, F2, and S2. In addition, other consideration factors may be applied to boost or demote the weighted scores, especially to break the tied weighted scores of F1 and S1. For example, if a user profile or previous user experience is in French, then F1 may be ranked prior to S1 in the response], the confidence scores for multiple languages are used to compute weighted scores, which are then used to determine ranking results, the weighted scores are boosted or demoted, resulting in changes in ranking results, i.e. French before Spanish. Modifying the weighted scores (which includes confidence scores) changes ranking between languages and because selection is based on these scores, modifying them updates the selected languages based on their associated confidence scores).
One of ordinary skill in the art would recognize that incorporating multiple sets of candidate languages, language recognition model, using language corpus to train the model; generating a confidence score of a language and updating language set based on confidence score of Sonntag to have a system which will have multiple candidates sets of languages, language recognition model and using multiple language corpus, generating a confidence score and updating the languages based on the confidence score to have a robust system. Datar, Sonntag are analogous arts because each art teaches processing languages.
Therefore, it would have been obvious to one of the ordinary skills in the art before the elective filing date to incorporate functionalities of Sonntag into the system of Datar to improve the accuracy and relevance of language selection for a given user query. Such modifications allow the system to prioritize languages that are more likely to be relevant to the user, thereby improving search results and overall system performance (Sonntag, [0021, The online game system employs a language analysis system to detect potential languages of a user query and to assist in efficient and reliable searching for results to formulate the response]).
Datar, Sonntag in combination do not explicitly teach identifying, by the data processing system and using the language recognition model, languages associated with one or more information resources associated with the one or more keywords; determining, by the data processing system, a second set of candidate languages based on the identified languages associated with the one or more information resources.
However, Datta teaches identifying, by the data processing system and using the language recognition model, languages associated with one or more information resources associated with the one or more keywords ([0024, consider a simple corpus in which only two languages are found: French and English. If "elephant" is a common-form entry in the synonyms map, then the variants "elephant", "elephant" and "elephant" would be associated with that entry as values, if these variants are found in the corpus… Each value also includes additional information: the languages of the documents in which instances of the variant appear, and the number of times the variant appears in that language. Continuing the example, "elephant" might be found in the corpus 90 times among documents recognized as being English and 300 times among documents recognized as being French], [0027, The language of each document in the training corpus is identified], [0032, The Naive Bayes models can be trained and tested using a large sample of electronic documents in various encodings and languages], multiple languages (French, English) associated with the documents (information resources) are determined by the using the Naïve Bayes models (language recognition model), wherein the documents are associated with keywords, for example, determining how many times the keyword elephant appear in the document (resources));
determining, by the data processing system, a second set of candidate languages based on the identified languages associated with the one or more information resources ([0043, The process 200 also removes any languages associated with a variant that have a frequency that does not exceed a pre-defined absolute threshold (step 230)], [0047, From each variant in the synonyms map, the process 200 removes any languages where the language's relative frequency does not meet a pre-defined relative threshold (e.g., 10%) (step 260). The same threshold applies to all variants and all languages], multiple languages associated with the documents (information resources) are determined, and languages that do not satisfy predetermined frequency thresholds are removed. The languages remaining after the removal/filtering are the “second set of candidate languages”).
One of ordinary skill in the art would recognize that incorporating identifying languages based on the keywords by the language recognition model and determining a second set of language to have a system which will process the retrieval of the resources faster. Datar, Sonntag, Datta are analogous arts because each art teaches processing languages.
Therefore, it would have been obvious to one of the ordinary skills in the art before the effective filing date to incorporate functionalities of Datta into the system of Datar/Sonntag to improve the accuracy and relevance of language selection for a given user query. Such modifications allow the system to prioritize languages that are more likely to be relevant to the user, thereby improving search results and overall system performance (Datta, [0056, adjustment of the query classifier can be iteratively repeated to further improve the accuracy.]).
With respect to claim 2, Datar, Sonntag, Datta in combination teaches the method of claim 1, Datar further teaches generating, by the data processing system, a first score for a first language of the plurality of languages based on a first number of occurrences of the first language in the browsing history of the account profile ([col. 4, lines 8-15, “users who speak Hindi may also be found, with some frequency, to speak English and to be interested in the content of a given page even though it is written in English. Interest in the content of the page by users having a Hindi language preference (indicated implicitly or explicitly) can be determined from aggregate statistics collected for the page from, for example, click logs, as described above”], [col. 8, lines “the other-language results section 604 can exclude results categorized (by a crawler application, for example) as having the same language as the user's preferred language, and include one or more highest ranked results in another language based on ranking scores calculated using language selection statistics”], the generation of scores of languages based on the users browsing histories which includes click logs, statistics (occurrences of the languages)).
Datar, Datta do not explicitly teach wherein the confidence scores are second confidence scores; a first confidence scores.
However, Sonntag teaches wherein the confidence scores are second confidence scores; first confidence scores ([0025, The machine learning model predicts candidate languages of the user query and outputs language confidence scores for the candidate languages], multiple candidate scores are calculated (first confidence score, second confidence score)).
One of ordinary skill in the art would recognize that incorporating generating a confidence score of a language and second confidence score of Sonntag into Datar to have multiple confidence score for languages. Datar, Sonntag, Datta are analogous arts because each art teaches processing languages.
Therefore, it would have been obvious to one of the ordinary skills in the art before the elective filing date to incorporate functionalities of Sonntag into the system of Datar/Datta to improve language selection by enabling more accurate identification of the most suitable language (Sonntag, [0021, The online game system employs a language analysis system to detect potential languages of a user query and to assist in efficient and reliable searching for results to formulate the response]).
With respect to claim 3, Datar, Sonntag, Datta in combination teaches the method of claim 2, Dater does not explicitly teach including, by the data processing system, the first language into the first set of candidate languages responsive to determining that the first confidence score for the first language is greater than a threshold score.
However, Sonntag teaches including, by the data processing system, the first language into the first set of candidate languages responsive to determining that the first confidence score for the first language is greater than a threshold score ([0031, Candidate languages for the word “pizza” may be Spanish and may also be English, and/or Italian. The language analysis system determines that all three candidate languages have confidence scores that meet a predefined threshold. The language analysis system 134 searches in individual databases associated with each of Spanish, English and Italian languages to find matching game related information related to the term “tycoon” and “pizza.” A match score is determined for each response match in the respective language-specific databases. The response score is based on the certainty of the database match. Search results to be included in a response is determined, based, at least in part, on weighted scores of the combined confidence score and response scores for the individual response matches.], determining the confidence score of a language is greater than a threshold to select that language to include in the set of languages).
One of ordinary skill in the art would recognize that incorporating including languages with exceeding confidence score into the language set of Sonntag into Datar, Datta to have a language with higher confidence score for languages. Datar, Sonntag, Datta are analogous arts because each art teaches processing languages.
Therefore, it would have been obvious to one of the ordinary skills in the art before the elective filing date to incorporate functionalities of Sonntag into the system of Datar, Datta to improve language selection by filtering less relevant languages to improve selection of relevant languages to have best search results (Sonntag, [0021, The online game system employs a language analysis system to detect potential languages of a user query and to assist in efficient and reliable searching for results to formulate the response.]).
With respect to claim 4, Datar, Sonntag, Datta in combination teaches the method of claim 1, Dater does not explicitly teach including, by the data processing system, a candidate language of the second set of candidates into the first set of candidate languages responsive to determining that a respective confidence score of the confidence scores for the at least some of the second set of candidate languages is greater than a threshold score.
However, Sonntag teaches including, by the data processing system, a candidate language of the second set of candidates into the first set of candidate languages responsive to determining that a respective confidence score of the confidence scores for the at least some of the second set of candidate languages is greater than a threshold score (0031, The language analysis system determines that all three candidate languages have confidence scores that meet a predefined threshold. The language analysis system 134 searches in individual databases associated with each of Spanish, English and Italian languages to find matching game related information related to the term “tycoon” and “pizza.”…. Search results to be included in a response is determined, based, at least in part, on weighted scores of the combined confidence score and response scores for the individual response matches], [0061, candidate languages may be ranked according to the respective candidate scores. Language-specific databases for the top designated number of ranked languages may be searched. In some implementations, language-specific databases of candidate languages that meet a threshold confidence score are used in the search], determining the confidence score of multiple languages is greater than a threshold to select that language to include in the set of languages).
One of ordinary skill in the art would recognize that incorporating including languages with exceeding confidence score into the language set of Sonntag into Datar, Datta to have a language with higher confidence score for languages. Datar, Sonntag, Datta are analogous arts because each art teaches processing languages.
Therefore, it would have been obvious to one of the ordinary skills in the art before the elective filing date to incorporate functionalities of Sonntag into the system of Datar, Datta to improve language selection by filtering less relevant languages to improve selection of relevant languages to have best search results (Sonntag, [0021, The online game system employs a language analysis system to detect potential languages of a user query and to assist in efficient and reliable searching for results to formulate the response]).
With respect to claim 5, Datar, Sonntag, Datta in combination teaches the method of claim 1, Dater further teaches identifying, by the data processing system, a first plurality of content items in a first language and a second plurality of content items in a second language ([col. 5, lines 50-55, “a language classifier 220 can aggregate the language preference statistics stored in the selection data repository 218 and use this aggregated data to update the content item index 210. The ranking engine 216 can then use these statistics and a language preference of a query submitted by a user to rank content items satisfying the query”], identifying plurality of content items in a first language and plurality of content items from the second languages);
and providing, by the data processing system to the client device, a content item selected from one of the first plurality of content items and the second plurality of content items, the content item in one of the first language or the second language (fig. 6, 7, [col. 8, lines “The results in section 602 can, for example, be top ranked results from a search for results in the user's preferred language”]).
Datar and Datta do not explicitly teach the updated first set of candidate languages; of the updated first set of candidate languages.
However, Sonntag teaches the updated first set of candidate languages; of the updated first set of candidate languages ([0060, A minimum threshold score for each candidate language may be returned by the machine learning model. Any language that returns a score over a minimum threshold is returned as a possible candidate language], [0061, candidate languages may be ranked according to the respective candidate scores. Language-specific databases for the top designated number of ranked languages may be searched. In some implementations, language-specific databases of candidate languages that meet a threshold confidence score are used in the search], the inclusion of languages exceeding the threshold and exclusion of others necessarily results in modification of the candidate language set, i.e. updating).
One of ordinary skill in the art would recognize that incorporating updated languages set of Sonntag into Datar to have updated language set. Datar, Sonntag, Datta are analogous arts because each art teaches processing languages.
Therefore, it would have been obvious to one of the ordinary skills in the art before the elective filing date to incorporate functionalities of Sonntag into the system of Datar, Datta to improve the accuracy of results to only have relevant results and excluding less likely candidates (Sonntag, [0021, The online game system employs a language analysis system to detect potential languages of a user query and to assist in efficient and reliable searching for results to formulate the response]).
With respect to claim 6, Datar, Sonntag, Datta in combination teaches the method of claim 1, Datar further teaches identifying, by the data processing system, a selection value for each content item of a first plurality of content items in a first language of the updated first set of candidate languages and a second plurality of content items in a second language of the updated first set of candidate languages ([col. 5, lines 25-30, “ The search engine 202 can use query and content information to generate a ranking score to rank the search results. The ranking score can be computed from information indicating how well each result satisfies or matches the query, from information indicating a level of query-independent quality of each result, or both”], items are selected based on the ranking (selection value), Sonntag teaches updated language sets (0061, candidate languages may be ranked according to the respective candidate scores. Language-specific databases for the top designated number of ranked languages may be searched. In some implementations, language-specific databases of candidate languages that meet a threshold confidence score are used in the search)); and
selecting, by the data processing system from the first plurality of content items and the second plurality of content items, a content item to provide to the client device in accordance to a content selection protocol, the content item in one of the first language or the second language (fig. 6, [col. 5, lines 25-30, “The search engine 202 can use query and content information to generate a ranking score to rank the search results. The ranking score can be computed from information indicating how well each result satisfies or matches the query, from information indicating a level of query-independent quality of each result, or both”], results are provided to the user’s client device).
With respect to claim 7, Datar, Sonntag, Datta in combination teach the method of claim 1, Datar further teaches identifying, by the data processing system, a third set of candidate languages from at least one of: (i) content in each information resource of a plurality of information resources identified in response to a request for content and a corresponding ranking of each information resource, (ii) a language configuration of an application executing on the client device, or (iii) one or more language settings associated with the account profile ([col. 6, lines 15-20, “ if a language preference of a query is Hindi, content items, in English, for example, that have user selection statistics indicating strong interest by other Hindi users can have increased scores (and therefore possibly increased rank positions) relative to content items with selection statistics that do not show this interest by Hindi users”], [col. 7, lines 60-63, “other user interface elements are in German because, in this example, John Doe previously indicated a German language preference”], multiple languages i.e. English, Hindi, German).
Datar, Datta do not explicitly teach updating, by the data processing system, the first set of candidate languages based on the third set of candidate languages.
However, Sonntag teaches and updating, by the data processing system, the first set of candidate languages based on the third set of candidate languages ([0060, A minimum threshold score for each candidate language may be returned by the machine learning model. Any language that returns a score over a minimum threshold is returned as a possible candidate language)], [0061, Language-specific databases for the top designated number of ranked languages may be searched. In some implementations, language-specific databases of candidate languages that meet a threshold confidence score are used in the search.], [0063, resulting boosted score of a particular language may reach a confidence threshold or ranking, even if the confidence score without the boost value fails to meet a threshold or ranking criteria. A demoted score may remove a particular language even if the confidence score without the demote value meets a threshold or ranking criteria. A language-specific database of a language with a boosted score that meets the searching criteria may be searched], languages are included/removed based on the confidence scores of other languages (updating based on other languages)).
One of ordinary skill in the art would recognize that incorporating updating languages of Sonntag into the system of Datar, Datta to have a system which will update the languages based on the other languages to have a robust system. Datar, Sonntag, Datta are analogous arts because each art teaches processing languages.
Therefore, it would have been obvious to one of the ordinary skills in the art before the elective filing date to incorporate functionalities of Sonntag into the system of Datar, Datta to improve language selection by enabling more accurate identification of the most suitable language (Sonntag, [0021, The online game system employs a language analysis system to detect potential languages of a user query and to assist in efficient and reliable searching for results to formulate the response]).
With respect to claim 8, Datar, Sonntag, Datta in combination teach the method of claim 1, Datar further teaches wherein the browsing history includes at least one of: a search query received from the client device, accessing of an information resource by the client device, and interaction with an element on information resource ([col. 6, lines 445-50, “ the selection data repository 218 includes a browsing history for users of the search engine, and browsing activity records can be used alone or in combination with selection records to generate language-specific statistics for a given content item. Users may be provided with an opportunity to opt in or opt out of the collection of browsing history or other features that may collect the personal information”], accessing users browsing histories which includes search query).
With respect to claim 9, Datar, Sonntag, Datta in combination teach the method of claim 1, Datar, Datta do not explicitly teach wherein the language recognition model is at least one of: (i) an artificial neural network, (ii) an n-gram model, (iii) a Bayesian network, (iv) a random forest model, (v) a support vector machine, or (vi) a decision tree model.
However, Sonntag teaches wherein the language recognition model is at least one of: (i) an artificial neural network, (ii) an n-gram model, (iii) a Bayesian network, (iv) a random forest model, (v) a support vector machine, or (vi) a decision tree model ([0007, receiving a user query and separating the user query into a plurality of n-grams. At least one candidate language is identified by applying a machine learning model to the plurality of n-grams of the user query], the languages recognition model is an n-gram model).
One of ordinary skill in the art would recognize that incorporating n-gram model of Sonntag to have a system which will have n-gram model to process languages. Datar, Sonntag, Datta are analogous arts because each art teaches processing languages.
Therefore, it would have been obvious to one of the ordinary skills in the art before the elective filing date to incorporate functionalities of Sonntag into the system of Datar, Datta to improve language processing and identification to find the best matching results (Sonntag, [0021, The online game system employs a language analysis system to detect potential languages of a user query and to assist in efficient and reliable searching for results to formulate the response.]).
Claim 11 encompasses the same scope of invention of claim 1, in additions of a system comprising: a data processing system having one or more processors coupled with memory, configured to: (Datar, [col. 10, lines 1-5, “Generally, a processor will receive instructions and data from a read-only memory or a random-access memory or both”]). Therefore, claim 11 is rejected on the same basis of rejection of claim 1.
Claim 12 is rejected on the same basis of rejection of claim 2.
Claim 13 is rejected on the same basis of rejection of claim 3.
Claim 14 is rejected on the same basis of rejection of claim 4.
Claim 15 is rejected on the same basis of rejection of claim 5.
Claim 16 is rejected on the same basis of rejection of claim 6.
Claim 17 is rejected on the same basis of rejection of claim 7.
Claim 18 is rejected on the same basis of rejection of claim 8.
Claim 19 is rejected on the same basis of rejection of claim 9.
Claim(s) 10, 20 is/are rejected under 35 U.S.C. 103 as being unpatentable over Datar et al. (US 8,375,025) and in view of Sonntag et al. (US 2021/0334299) and in view of Datta et al. (US 2007/0288450) and in view of Moon et al. (US 2016/0026913).
With respect to claim 10, Datar, Sonntag, Data in combination teaches the method of claim 1, Datar does not explicitly teach applying, by the data processing system, each of the corpuses of text for each language of the plurality of languages to the training dataset to generate a set of results corresponding to result languages of the plurality of languages; generating, by the data processing system, a result error by comparing each of the result languages to a labeled language for each of the corpuses; and modifying, by the data processing system, one or more weights of the language recognition model based on the result error.
However, Sonntag teaches wherein training the language recognition model includes: applying, by the data processing system, each of the corpuses of text for each language of the plurality of languages to the training dataset to generate a set of results corresponding to result languages of the plurality of languages ([0004, The machine learning model is trained with at least one multilingual text corpus and game-related data. A respective language confidence score for each of the at least one candidate language is determined by applying the machine learning model], [0051, the machine learning model analyzes the training text data and generates predicted language as output data], [0050, In block 302, the processed training data may be fed into an untrained machine learning model without the associated language labels for supervised training], [0052, In block 306, the current predicted language labels are compared with language labels that are associated with the training data prior to the training, such as the language labels from the pre-processing of the training data.], [0076, A listing of game information for a particular language may be validated by reference to supplemental language information], the language model is trained according to a training datasets which includes corpuses from multiple languages with a particular language label (for each language), to generate predicted language as output data (output as results));
generating, by the data processing system, a result error by comparing each of the result languages to a labeled language for each of the corpuses by comparing each of the result languages to a labeled language for each of the corpuses ([0010, Retraining of the machine learning model may be conducted with discrepancy information between the current predicted language labels and the associated language labels], [0052, Discrepancy information is generated that is indicative of the difference between the predicted labels and the previously associated labels. In decision block 308, it may be determined if the discrepancy information meets a pre-defined threshold for the accuracy of the predicted language labels], generating discrepancies (result error data) by comparing predicated output label (result) with a previously associated label (labeled language)),
One of ordinary skill in the art would recognize that incorporating training the language recognition model based on corpus of text from multiple languages to generate results corresponding to plurality of languages and generating results error by comparing languages of Sonntag into the invention of Datar, Data to train the language learning model using corpus of languages and also to generate result error or languages to identify most appropriate languages faster. Datar, Sonntag, Data are analogous arts because each art teaches processing languages.
Therefore, it would have been obvious to one of the ordinary skills in the art before the elective filing date to incorporate functionalities of Sonntag into the invention of Datar, Data to understand diverse linguistic patterns, make it more robust and adaptable across various language contexts which will improve models performance in multilingual applications, such as translation, sentiment analysis and natural language understanding (Sonntag, [0021, The online game system employs a language analysis system to detect potential languages of a user query and to assist in efficient and reliable searching for results to formulate the response]).
Datar, Sonntag, Data do not in combination explicitly teach modifying, by the data processing system, one or more weights of the language recognition model based on the result error.
However, Moon teaches modifying, by the data processing system, one or more weights of the language recognition model based on the result error ([0051, The neural network trainer 120 may compute an error by comparing a desired expectation value for the training data and an output value generated from the output layer of the neural network, and may adjust the connection weight applied to the recognition model of the neural network to reduce the error], the model is modified based on the error and Sonntag teaches language recognition model in paragraph [0030, The pre-processing enables the machine learning model 136 to recognize languages from a small number of input words in the user quay]).
One of ordinary skill in the art would recognize that incorporating modifying weights of the model based on the error of Moon into the invention of Datar, Sonntag, Datta to update model weights based on error. Datar/Sonntag/Datta/Moon are analogous arts because each art teaches processing data.
Therefore, it would have been obvious to one of the ordinary skills in the art before the elective filing date to incorporate functionalities of Moon into the invention of Datar/Sonntag/Datta to optimize the language model, thereby, increasing confidence reliability and improving classification performance across candidate languages (Moon, [0088, the data processing apparatus 600 may continuously improve a recognition performance by continuously performing training of sequential data failing to be recognized]).
Claim 20 is rejected on the same basis of rejection of claim 10.
Prior art Made of Record
The prior arts made of record and not relied upon are considered pertinent to the applicant’s disclosure.
Sanjekar et al (US 2019/0294685) teaches determining languages in [0040].
Poiesz et al. (US 2015/0347378) teaches determining languages in paragraphs [0020-022].
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 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 FATIMA P MINA whose telephone number is (571)270-3556. The examiner can normally be reached Monday - Friday 9:00 am - 5:00 pm.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Ann Lo can be reached at 571-272-9767. 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.
/FATIMA P MINA/ Examiner, Art Unit 2159
/ANN J LO/ Supervisory Patent Examiner, Art Unit 2159
Prosecution Insights