Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Acknowledgement
Acknowledgement is made of applicant’s amendment made on 03/02/2026. Applicant’s submission filed has been entered and made of record.
Status of the Claims
Claims 1-14 are pending.
Response to Applicant’s Arguments
In response to “The quoted limitations cannot be performed mentally. For example: " A human being cannot traverse a graph "based on an information simplification topology graph prebuilt based on the graph data", because such a topology graph is a specialized computer-constructed data structure comprising node identifiers stored in memory. " A human cannot perform the required searching "along an edge in the graph data" without materializing intermediate nodes, because such edge-indexed memory structures do not exist in the human mind”.
However, claims 1, 6, and 10 do not recite any particular computer constructed data structure to implement the processing of to be processed text mentioned above that goes beyond human mental judgment.
Rather, claims 1, 6, and 10 are entirely focused on obtaining a prediction result, which corresponds to mental processes that are non-statutory.
In response to “In this way, a natural language understanding method with a stronger generalization capability is provided. In addition, the prediction result indicates how to split the entire to-be-processed text and further includes at least one label corresponding to the to-be-processed text, that is, the prediction result carries more abundant information, therefore helping improve accuracy of a process of understanding the intention of the to-be-processed text” and “Moreover, the amended claim further recites wherein "the prediction result indicates a target splitting manner corresponding to the to-be-processed text and the target splitting manner is obtained based on degrees of matching between the plurality of target character sets and the plurality of first labels." As explicitly disclosed in the Speciation, "According to this solution, at least one first label capable of more accurately indicating the intention of the to-be-processed text can be obtained."”.
According to an example disclosed by the specification US 2024/0220730 A1 at ¶134:
“As shown in FIG. 7, the to-be-processed text is “Make a query of the mobile number for a specific place”, and the prediction result of the to-be-processed text shows that the to-be-processed text is split into four first character sets: “Make a query of”, “the mobile number”, “for”, and “a specific place”. As shown in FIG. 7, the prediction result of the to-be-processed text further includes five first labels: “Home location query”, “Query number”, “Home location”, “Query”, and “Phone number”. The label “Home location query” represents semantics of the entire to-be-processed text, the label “Query number” represents semantics of “Make a query of the mobile number”, the label “Home location” represents semantics of “a specific place”, the label “Query” represents semantics of “Make a query of”, and the label “Phone number” represents semantics of “the mobile number”, so that an intention of the entire to-be-processed text can be understood based on the five first labels”.
In another example disclosed by the specification US 2024/0220730 A1 at ¶138:
“if the to-be-processed text is “Make a phone call to Xiao Ming”, a splitting manner corresponding to the to-be-processed text may be “to Xiao Ming” and “Make a phone call”; another splitting manner may be “to”, “Xiao Ming”, and “Make a phone call”; still another splitting manner may be “to Xiao Ming”, “Make”, and “a phone call”; and yet another splitting manner may be “Make to Xiao Ming” and “a phone call”, or the like”.
In other words, improving the labeling of “Make a query of the mobile number for a specific place” and making prediction or understanding the meaning of “Make a phone call to Xiao Ming” are entirely mental judgment processes, which are non-statutory.
Patent eligibility requires a specifically asserted improvement in computer capabilities or functionalities.
For example, the specification described a specifically asserted machine learning model structure for natural language processing in artificial intelligence field comprising:
“The first model may include an encoder and a decoder. For example, the encoder may use a transformer (transformer) structure, and the decoder may specifically use a multi-layer perceptron (multi-layer perceptron, MLP)”; the specification US 2024/0220730 A1 at ¶146;
“Specifically, the training device may input the first character set and the at least one second label (or the null label) into the encoder, to perform feature extraction by using the encoder, to generate a vector representation corresponding to the first character set; and the training device inputs the vector representation corresponding to the first character set into the decoder, to generate the plurality of second scores by using the decoder”; the specification US 2024/0220730 A1 at ¶147; and
“In a semantic understanding phase, the training device obtains a target character string from the to-be-processed text, and obtains, from the first data set, at least one second label that matches the target character string; the training device generates, based on the target character string and the at least one second label by using an encoder in a target model, a vector representation of the target character string; and the training device generates, based on the vector representation of the target character string, a belonging relationship between labels at different levels in the first data set, and all third labels in the first data set by using a decoder, a prediction result corresponding to the to-be-processed text”; the specification US 2024/0220730 A1 at ¶164.
Therefore, if claims 1, 6, and 10 are amended to require (1) a specifically asserted machine learning model structure and (2) how the structure is applied in a natural language processing of the to be processed text to determine intention of the to be processed text, then such specifically asserted / claimed machine learning model implemented natural language processing would go beyond human mental steps of making predictions.
In response to “As shown, the cited portion of Zhou merely discloses performing word segmentation processing and there are multiple word segmentation method. But Zhou has not been shown the "a target splitting manner corresponding to the to-be-processed text" is indicated in "a prediction result" resulting from "processing the to-be-processed text by using a target model," let alone that "the target splitting manner is obtained based on degrees of matching between the plurality of target character sets and the plurality of first labels" as recited in claim 1” and “Indeed, Zhang has not been shown to disclose determining a splitting manner of the to-be- processed text, but rather taking an already segmented result as an input to an intent recognition model. See Zhang, [0046]-[0047] ("In S301, a to-be-recognized text is acquired." "In S302, word segmentation results of the to-be-recognized text are inputted to an intent recognition model, and a first intent result and a second intent result of the to-be-recognized text are obtained according to an output result of the intent recognition model.") Moreover, Zhang's "word segmentation results of the to-be-recognized text" (to the extent it is mapped to "the target splitting manner") is an input to "an intent recognition model," rather than an output from "processing the to-be-processed text by using a target model" comprised in the "prediction result." As such, Zhang has not been shown to disclose at least "processing the to-be-processed text by using a target model to obtain a prediction result, wherein...the prediction result indicates a target splitting manner corresponding to the to-be-processed text and the target splitting manner is obtained based on degrees of matching between the plurality of target character sets and the plurality of first labels" as recited in claim 1”.
In view of the aforementioned amendments, rejection under Zhang and Zhou have been withdrawn. Upon further search and consideration, please see details of a new combination of references set forth below.
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 1-14 are rejected under 35 USC 101 as directing toward non-statutory subject matter.
Claims 1 and 6 recite a method (“process”). Claim 10 recites a data processing apparatus comprising a processor and memory (“machine”).
To distinguish ineligible claims that merely recite a judicial exception from eligible claims that require an implementation of judicial exception, the Supreme Court uses a two-step framework: Step One (Step 2A), determine whether the claims at issue are directed to one of those patent-ineligible concepts; and Step Two (Step 2B), if so, ask “what else is there in the claims?” to determine whether the additional elements transform the nature of the claim into a patent eligible application. Alice Corp. Pty. Ltd. v. CLS Bank Int’l., 134 S. Ct. 2347, 2355 (2014).
Step One (Step 2A) is a two prong test that requires the determination of whether the claims at issue are directed to an enumerated patent ineligible concept. See MPEP 2106.04.
Specifically, Step 2A Prong (1) requires the determination of the specific limitations in the claim under examination (individually or in combination) that the examiner believes recites an abstract idea and determining whether the identified limitations falls within the subject matter groupings of abstract ideas enumerated. See MPEP 2106.04(a).
The enumerated patent ineligible concepts comprising:
(a) Mathematical Concepts – mathematical relationships, mathematical formulas or equations, mathematical calculations;
(b) Certain methods of organizing human activity – fundamental economic principles / practices (including hedging, insurance, mitigating risk); commercial or legal interactions (including agreements in the form of contracts; legal obligations; advertising, marketing or sales activities or behaviors; business relations); managing personal behavior or relationships or interactions between people (including social activities, teaching, and following rules / instructions) and
(c) Mental processes – concepts performed in the human mind (including an observation, evaluation, judgment, opinion). See MPEP 2106.04(a).
If the claim recites an enumerated patent ineligible concept, then Prong (2) of Step One (Step 2A) requires the determination of whether the claim integrates the patent ineligible concept into a practical application. Individually and in combination, identifying whether there are any additional elements recited in the claim beyond the judicial exceptions and evaluating those additional elements to determine whether they integrate the exception into a practical application, using one or more of the considerations laid out by the Supreme Court and the Federal Circuit. See MPEP 2106.04(d).
Under Step Two (Step 2B), if the claim does not integrate the ineligible concept into a practical application and therefore directed to a judicial exception, evaluate whether the claim provides an inventive concept by determining whether there are additional elements, individually and in ordered combination, amount to significantly more than the exception itself. See MPEP 2106.04.
Step 2A Prong (1)
The “directed to” inquiry does not ask whether the claims involve a patent ineligible concept but, considered in light of the specification, whether the claim as a whole is directed to excluded subject matter or directed to an improvement to computer functionality. Enfish L.L.C. v. Microsoft Corp., 822 F.3d 1327, 1335 (Fed. Cir. 2016).
Therefore, Prong (1) of Step 2A requires identifying specific limitations in the claims that recites (“describes” or “set forth”) an abstract idea and determine whether the identified limitations falls within the subject matter groupings of abstract ideas enumerated. See MPEP 2106.04 (“Thus, it is sufficient for this analysis for the examiner to identify that the claimed concept (the specific claim limitation(s) that the examiner believes may recite an exception) aligns with at least one judicial exception”).
In particular, MPEP 2106.04(a)(2) states “a claim that recites a mathematical calculation, when the claim is given its broadest reasonable interpretation in light of the specification, will be considered as falling within the "mathematical concepts" grouping. A mathematical calculation is a mathematical operation (such as multiplication) or an act of calculating using mathematical methods to determine a variable or number, e.g., performing an arithmetic operation such as exponentiation”.
Under Prong (1), claim 10 recites an apparatus for text data processing, wherein the apparatus comprises: one or more processors; and one or more memories coupled to the one or more processors and storing programming instructions for execution by the one or more processors to:
(1) obtain a to-be-processed text, wherein the to-be-processed text comprises a plurality of characters; and
(2) process the to-be-processed text by using a target model to obtain a prediction result, wherein the target model is a machine learning model for natural language processing in artificial intelligence field, the prediction result indicates to split the to-be-processed text into a plurality of target character sets, each target character set of the plurality of target character sets comprises at least one character, the prediction result further comprises a plurality of first labels, each first label of the plurality of first labels indicates semantics of a respective target character set of the plurality of target character sets, and the plurality of first labels are used to determine an intention of the to-be-processed text, and the prediction result indicates a target splitting manner corresponding to the to-be-processed text and the target splitting manner is obtained based on degrees of matching between the plurality of target character sets and the plurality of first labels.
Claim 1 recites a corresponding method.
Claim 6 recites a method for neural-network training, wherein the method comprises:
(2) processing a to-be-processed text by using a target model to obtain a prediction result, wherein the target model is a machine learning model for natural language processing in artificial intelligence field, the to-be-processed text comprises a plurality of characters, the prediction result indicates to split the to-be-processed text into a plurality of target character sets, each target character set of the plurality of target character sets comprises at least one character, the prediction result further comprises a plurality of first labels, each first label of the plurality of first labels indicates semantics of a respective target character set of the plurality of target character sets, and the plurality of first labels are used to determine a predicted intention of the to-be-processed text, and the prediction result indicates a target splitting manner corresponding to the to-be- processed text and the target splitting manner is obtained based on degrees of matching between the plurality of target character sets and the plurality of first labels; and
(3) training the target model according to a target loss function to obtain a trained target model, wherein the target loss function indicates a similarity between the prediction result and an expected result corresponding to the to-be-processed text, the expected result corresponding to the to-be-processed text indicates to split the to-be-processed text into a plurality of second character sets, each second character set comprises at least one character, the expected result corresponding to the to-be-processed text further comprises a plurality of expected labels, one expected label indicates semantics of one second character set, and the plurality of expected labels are used to determine a correct intention of the to-be-processed text.
Individually, step (1) for obtaining to be processed text corresponds to collecting information. Collecting information, including when limited to particular content, is within the realm of abstract ideas. Electric Power Grp., L.L.C. v. Alstom SA, 830 F.3d 1350, 1353 (Fed. Cir. 2016).
Individually, step (2) for using a target model to process the to be processed text to obtain a prediction result corresponds to making evaluation or judgment. In view of the specification US 2024/0220730 A1 at ¶79: “The training device 220 generates a target model/rule 201, and performs iterative training on the target model/rule 201 by using the training data set in the database 230, to obtain a trained target model/rule 201. The trained target model/rule 201 may also be referred to as a mature target model/rule 201. Further, the target model/rule 201 may be specifically implemented by using a neural network model or a model of a non-neural network type”.
Under the broadest reasonable interpretation, step (2) can be accomplished by a person performing mental step to make a prediction. Analyzing information by steps people go through in their minds, or by mathematical algorithms, without more, are treated as essentially mental process within the abstract-idea category. Electric Power Grp., 830 F.3d at 1354. Therefore, step (2) is essentially a mental process.
Individually, step (3) corresponds to applying mathematical weights based on target keyword and the synonym group. In view of the specification US 2024/0220730 A1 at ¶117: “Further, the training device may input a candidate training text into a first model, to compute, by using the first model, a perplexity loss (perplexity loss) corresponding to the candidate training text, and then obtain a third score that is output by the first model and that corresponds to the candidate training text. For example, the training device generates six candidate training texts based on one target data subset, and obtains a third score for each candidate training text. Details are shown in the following table”:
PNG
media_image1.png
173
457
media_image1.png
Greyscale
Under the broadest reasonable interpretation, step (3) can be accomplished by a person performing mental / mathematical calculations to adjust a math model. Therefore, step (3) is essentially a mental process.
In ordered combination, steps (1)-(3) correspond to collecting information limited to particular content (to be processed text with characters), analyzing information by steps people go through in their minds to make intent predictions, and analyzing information by mathematical algorithm to mathematically calculate target loss functions to adjust a math model.
Thus, claims 1, 6, and 10 described patent ineligible subject matter enumerated under category (a) Mathematical Concepts – mathematical relationships, mathematical formulas or equations (target model as a math model), mathematical calculations (calculating loss functions) and (c) Mental processes – concepts performed in the human mind (including an observation, evaluation, judgment (making intent predictions), opinion).
Step 2A Prong (2).
Under Prong (2) of Step 2A, the goal is to determine whether the claim is directed to the recited exception by evaluating whether the claim as a whole integrates the recited judicial exception into a practical application of the exception. See MPEP 2106.04II(A).
In particular, evaluating integration into a practical application requires identifying whether there are any additional elements recited in the claim beyond the judicial exception and evaluating those additional elements, individually and in combination, to determine whether they integrate the exception into a practical application, using one or more of the considerations laid out by the Supreme Court and the Federal Circuit (“CAFC”). See MPEP 2106.04(d).
The Supreme Court held that when a claim containing a mathematical formula (i.e., an abstract idea) implements or applies that math formula / abstract idea in a structure or process which, when considered as a whole, is performing a function which the patent laws were designed to protect (e. g., transforming or reducing an article to a different state or thing), then the claim satisfies the requirements of §101. Diamond v. Diehr, 450 U.S. 175, 192 (1981); See MPEP 2106.04(d)I (“Implementing a judicial exception with, or using a judicial exception in conjunction with, a particular machine or manufacture that is integral to the claim, as discussed in MPEP 2106.05(b)”). See also Benson, 409 U.S. at 70 (“Transformation and reduction of an article "to a different state or thing" is the clue to the patentability of a process claim that does not include particular machines”).
In one example, the CAFC applied Alice inquiry to ask whether the focus of the claims is on the specific asserted improvement in computer capabilities (i.e., the self-referential table for a computer database) or instead, on a process that qualifies as an abstract idea for which computers are invoked merely as a tool. Enfish L.L.C. v. Microsoft Corp., 822 F.3d 1327, 1335-36 (Fed. Cir. 2016).
In Enfish, the claims were specifically directed to a self-referential table for a computer database. Id. at 1337. In particular, the claim language required a four step algorithm specifically directed to a self-referential table for a computer database that improves upon prior art information search and retrieval systems by employing a flexible, self-referential table to store data. Id. at 1336-37. CAFC determined that the plain focus of the claims was on an improvement to computer functionality itself (i.e., the self-referential table for a computer database), not on economic or other tasks for which a computer is used in its ordinary capacity. Id at 1335-36.
Therefore, the focus of the claims is on a specific asserted improvement in computer capabilities (i.e., the self-referential table for a computer database), not on economic or other tasks for which a computer is used in its ordinary capacity. Id. at 1336. See also MPEP 2106.04(d)I (“an improvement in the functioning of a computer or an improvement to other technology or technical field, as discussed in MPEP 2106.04(d)(1) and 2106.05(a)”).
In another example, the Supreme Court looked to how the claims used that equation in a process designed to solve a technological problem in conventional industry practice. McRO, Inc. v. Bandai Namco Games America Inc., 837 F.3d 1299, 1312 (Fed. Cir. 2016).
Specifically, in Diehr, the claims involved a method for curing rubber by using Arrhenius equation to constantly measure actual temperature inside a mold and feeding the temperature measurements into a computer to repeatedly recalculate the cure time to open the press. Diehr, 450 U.S. at 178-79. Since the Supreme Court viewed the claims not as an attempt to patent a mathematical formula, but to an industrial process for molding of rubber products, the claims were statutory. Id. at 192-93.
The key here, as noted by the CAFC, is that the Supreme Court in Diehr looked to how the claims "used that equation in a process designed to solve a technological problem in `conventional industry practice.'" McRO, 837 F.3d at 1312. When looked at as a whole, "the claims in Diehr were patent eligible because they improved an existing technological process, not because they were implemented on a computer." Id. at 1312-13.
In McRO, the CAFC noted that prior art method of generating (i.e., calculating) morph weight set with values between “0” and “1” for computer animation of facial expressions are manually determined. McRO, 837 F.3d at 1304-5. The claimed improvement in McRO allows computers to produce “accurate and realistic lip synchronization and facial expressions in animated characters” that previously could only be produced by human animators through the automated use of rules, rather than artists, to set the morph weights and transitions between phonemes. Id. at 1313.
Specifically, the claims are directed to the incorporation of claimed rules, not the use of the computer that improved existing technological process by allowing automation of further tasks that goes beyond merely organizing existing information into a new form. Id. at 1314-15.
In other words, the claimed process uses a combined order of specific rules that renders information into a specific format that is then used and applied to create a sequence of synchronized, animated characters that prevent pre-emption of all processes for achieving automated lip-synchronization of 3-D characters. Id. at 1315. Therefore, the CAFC held that the ordered combination of claimed steps, using unconventional rules that relate sub-sequences of phonemes, timing, and morph weight sets is patent eligible. Id. at 1302-3.
Further, in USPTO’s Memo on 2024 Updated Guidance on AI and Subject Matter Eligibility issued July 16, 2024 describing example 48 on pp. 14-15 regarding a method to separate speech signals from different sources to recognize human speech command from background noise by using a deep neural network (DNN) to promote separation of the features during clustering. See p. 15, ¶2. Specifically, the DNN learns high level feature representations of the signal x by mapping the feature representations to the embedding space comprising the DNN converting feature representations Xt, obtained from spectrograms St and corresponding feature matrices FMt, into multi-dimensional embedding vectors V and assigning the embedding vectors V to TF bins as a global function of the input signal (V= fθ(X), where fθ represents a function of the DNN). See p. 15, ¶5.
According the Memo under Step 2A, Prong One, a claim comprising a step of using a deep neural network (DNN) to determine embedding vectors V using the formula V= fθ(X), where fθ represents a function of the DNN describes a mathematical calculation and therefore the claim “set forth” or “describes” a judicial exception. p. 19, ¶5.
Under Step 2A, Prong Two, since there is no detail about a particular DNN or how the DNN operates to derive the embedding vectors other than that it is being used to determine the embedding vectors, the DNN is used to generally apply the abstract idea of performing mathematical calculation using recited mathematical equation without placing any limitation on how the DNN operates to derive the embedding vectors as a function of the input signal. p. 20, ¶2.
In particular, the disclosure identifies a technical problem encountered in the field of speech separation and provides an improvement over existing speech separation methods by determining embedding vectors as a function of the input signal, partitioning those vectors into clusters, and synthesizing a reconstructed mixed speech signal based on these clusters. p. 20, ¶3. The claim, however, only requires determining the embedding vectors and therefore does not reflect the improvement discussed in the disclosure. Id.
In the instant application, claims 1, 6, and 10 set forth steps (1)-(3) correspond to collecting information limited to particular content (to be processed text with characters), analyzing information by steps people go through in their minds to make intent predictions, and analyzing information by mathematical algorithm to mathematically calculate target loss functions to adjust a machine learning target model to make intent predictions.
Unlike the specifically asserted self-referential table described by a four step algorithm in Enfish, claims 1, 6, and 10 broadly asserted that the target model is a machine learning model for natural language processing in artificial intelligence field. Such limitation does not specifically assert a particular structure for the target model or describe a particularly improved machine learning model as a structure for the target language model. The only assertion is that the target model is a machine learning model, which amounted to a broad recitation that it is tied to a computer or machine.
Unlike the particular means or method for applying the Arrhenius equation (i.e., an abstract idea) in a particular industrial process for curing rubber in Diehr, claims 1, 6, and 10 do not recite any particular technological or industrial application of the target model (i.e., a math model like the Arrhenius equation) because determining an intention of the to be processed text is a mental step like making a judgment or evaluation.
Unlike the technical incorporation of particular rules to the generated (i.e., calculated) morph weights to automate the “accurate and realistic lip synchronization and facial expressions in animated characters” in McRO, claims 1, 6, and 10 do not describe any particular technological application that would go beyond mathematical analysis to make a mental judgment or evaluation of the intention of the to be processed text or to train a new math model / target model.
Even interpreting the target model as neural network model per specification US 2024/0220730 A1 at ¶79, much like the claims in example 48 that merely used the DNN to calculate embedding vectors using a math formula without limitations on how the DNN is configured to synthesize a reconstructed speech signal, claims 1, 6, and 10 focused on generating the neural network model without limitations on how the neural network model is structured (e.g., US 2024/0220730 A1 at ¶146, model includes an encoder with transformer structure and decoder) and how such structure is applied to make natural language processing of to be processed text to determine intention of the to be processed text.
Finally, to the extent that claim 10 recited processor, attending software (i.e., instructions), and memory, the Supreme Court held that mere recitation of a generic computer cannot transform a patent-ineligible abstract idea into a patent-eligible invention. Alice, 134 S. Ct. at 2358.
For example, in Alice, the Supreme Court held that data processing systems with communication controller, data storage unit, and transmission units were purely functional and generic because nearly every computer will include a "communications controller" and "data storage unit" capable of performing the basic calculation, storage, and transmission functions and such recitation of hardware failed to offer any meaningful limitation beyond generally linking the use of a method to a particular technological environment. Id. at 2360. See MPEP 2106.04(d)I (“Generally linking the use of a judicial exception to a particular technological environment or field of use, as discussed in MPEP 2106.05(h)”). Neither stating an abstract idea while adding the words “apply it” nor limiting the use of an abstract idea to a particular technological environment is enough for patent eligibility. Id. at 2350.
Much like the data processing systems with data storage unit performing basic calculations in Alice, the recitation of processor, program instructions, and memory in claim 10 are purely functional and generic that failed to offer any meaningful limitation beyond generally linking the claims to computers.
Therefore, claims 1, 6, and 10 are directed to collecting information limited to particular content (to be processed text with characters), analyzing information by steps people go through in their minds to make intent predictions, and analyzing information by mathematical algorithm to mathematically calculate target loss functions to adjust a math model.
Step 2B Inventive Concept.
The Guideline stated that if the additional elements do not integrate the exception into a practical application, then the claim is directed to the recited judicial exception, and requires further analysis under Step 2B where it may still be eligible if it amounts to an “inventive concept”. See MPEP 2106.04IIA and MPEP 2106.05.
Further, an inventive concept can be found in the non-conventional and non-generic arrangement of known conventional pieces. BASCOM Global Internet Servs. v. AT&T Mobility, 827, F3d 1341, 1350 (Fed. Cir. 2016).
In BASCOM, the CAFC held that filtering content is an abstract idea because it is a longstanding, well-known method of organizing human behavior similar to concepts previously found to be abstract. BASCOM, 827 F.3d at 1348. However, the CAFC determined that the claims did not merely recite filtering content along with the requirement to perform it on the internet or on a set of generic computer components, nor did the claims preempt all ways of filtering content on the internet. Id. at 1350.
Rather, the inventive concept described and claimed was the installation of a filtering tool at a specific location, remote from the end-users, with customizable filtering features specific to each end user that gives the filtering tool both the benefits of a filter on a local computer and the benefits of a filter on an internet service provider “ISP” server. Id. By taking a prior art filter solution (one size fits all filter at internet service provider “ISP” server) and making it more dynamic and efficient (providing individualized filtering at the ISP server), the claimed invention improves the performance of the computer system itself. Id. at 1351.
On the other hand, implementation via computers does not offer a meaningful limitation beyond generally linking the use of an abstract idea to a particular technological environment. Alice, 134 S. Ct. at 2360 (“Nearly every computer will include a “communications controller” and “data storage unit” capable of performing the basic calculation, storage, and transmission functions required by the method claims”). Intellectual Ventures I L.L.C. v. Capital One Bank, 792 F.3d 1363, 1370-71 (Fed. Cir. 2015) (“Steps that do nothing more than spell out what it means to “apply it on a computer” cannot confer patent-eligibility).
Similarly, limiting an abstract idea to one field of use do not convert otherwise ineligible concept into an inventive concept. Intellectual Ventures I L.L.C. v. Erie Indem. Co., 850 F.3d 1315, 1328 (Fed. Cir. 2017). Neither does adding computer functionality to increase the speed or efficiency of the process confer patent eligibility on an otherwise abstract idea. Intellectual Ventures I, 792 F.3d at 1367 (citing Bancorp Servs., LLC v. Sun Life Insurance Co. of Can., 687 F.3d 1266, 1278 (Fed. Cir. 2012) (“The fact that the required calculations could be performed more efficiently via a computer does not materially alter the patent eligibility of the claimed subject matter”)).
Individually, in the instant application, claims 1, 6, and 10 set forth steps (1)-(3) corresponding to collecting information limited to particular content (to be processed text with characters), analyzing information by steps people go through in their minds to make intent predictions, and analyzing information by mathematical algorithm to mathematically calculate target loss functions to adjust a math model.
Claim 10 further requires computer processor, memory storing program instructions for implementing steps (1)-(2).
Such individual recitation of generic computer components (processor, software / program instructions) are purely functional and generic because nearly every computer will include such processor and data storage unit capable of performing basic calculation necessary for step (2) to make an evaluation / judgment / prediction and step (3) to mathematically calculate loss function values to generate a math model.
Further, individually, asserting that the target model is a machine learning model for natural language processing in artificial intelligence field broadly limits claims 1, 6, and 10’s target model to the field of use of natural language processing. However, limiting an abstract idea to one field of use do not convert otherwise ineligible concept into an inventive concept.
As an ordered combination, unlike BASCOM that describes an unconventional combination of a conventional ISP server with a customized filter specific to each user that is remote from end-users to provide both the benefits of a filter on a conventional local computer and the benefits of a filter on the conventional ISP server, using target model that is either a set of rules, math model, or neural network with software program instruction + processor to make intention prediction of to be processed text do not involve a unconventional combination of conventional pieces because the combination amounts to “apply it on a computer”, which cannot convert otherwise ineligible concept into an inventive concept.
To the extent that implementing target model in the field of computers results in reduction in memory requirement and computational requirement, merely adding computer functionality to increase the speed or efficiency does not confer patent eligibility on an otherwise abstract idea.
Dependent claims 2, 7, and 11 do not recite any specifically asserted technological structure or technological application because intention prediction labeling of to be processed text can be accomplished mentally.
Dependent claims 3-5 and 12-14 do not recite any specifically asserted technological structure or technological application because splitting to be processed text to obtain prediction result can be accomplished mentally.
Dependent claim 8 does not recite any specifically asserted technological structure or technological application because making determination of the to be processed text and the expected result to determine a third score can be accomplished mathematically and mentally.
Therefore, claims 1-14 are not eligible for a patent.
Claim Rejections - 35 USC § 103
The following is a quotation of the appropriate paragraphs of 35 U.S.C. 103 that form the basis for the rejections under this section made in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
Claims 1, 3, 6, 8-10, and 12 are rejected under 35 USC 103(a) as being unpatentable over Zhang et al. (US 2023/0004798 A1) in view of Ramaswamy et al. (US 2014/0156268 A1).
Regarding Claims 1 and 10, Zhang discloses an apparatus for text data processing (Figs. 6-7, Intent Recognition Apparatus 600 being implemented by device 700), wherein the apparatus comprises:
one or more processors (¶76, computing unit 701 / CPU); and
one or more memories coupled to the one or more processors and storing programming instructions for execution by the one or more processors (¶74 and ¶76, computer program stored in storage unit 708 as computer software program for intent recognition methods) to:
obtain a to-be-processed text, wherein the to-be-processed text comprises a plurality of characters (¶46, step S301, acquire to be recognized text; e.g., ¶51, “Open the navigation app and take the highway”); and
process the to-be-processed text by using a target model to obtain a prediction result (¶47, S302, input word segmentation results of the to be recognized text into an intent recognition model to output intent results), wherein the target model is a machine learning model for natural language processing in artificial intelligence field (¶51 and Fig. 4, intent recognition model being a trained neural network comprising a first recognition layer outputs scores between word segmentation results in the to be recognized text and candidate intents and a second recognition layer processing word segmentation results to obtain second intent results), the prediction result indicates to split the to-be-processed text into a plurality of target character sets, each target character set of the plurality of target character sets comprises at least one character (¶47, input word segmentation results of the to be processed text into the intent recognition model; ¶51, input word segmentation results of “Open the navigation app and take the highway” comprising “open”, “navigation app”, “take”, and “highway” into the intent recognition model to obtain first intent results through the first recognition layer “NAVI” and “HIGHWAY”), the prediction result further comprises a plurality of first labels, each first label of the plurality of first labels indicates semantics of a respective target character set of the plurality of target character sets (¶51 and Fig. 4, for semantic vectors h1 “open”, h2 “navigation app”, h3 “take”, and h4 “highway”, intent recognition model outputs first intent results “NAVI” and “HIGHWAY”), the plurality of first labels are used to determine an intention of the to-be-processed text (¶51, first recognition layer of the intent recognition model outputs scores between the word segmentation results in the to be recognized text and the candidate intents; per ¶48, that is, intent recognition is performed on the to be recognized text by using the intent recognition model), and the prediction result indicates a target splitting manner corresponding to the to-be-processed text and degrees of matching between the plurality of target character sets and the plurality of first labels (¶51, to be processed text “Open the navigation app and take the highway” were segmented into first semantic vectors h1, h2, h3, and h4 and the first recognition layer outputted a score matrix indicating scores between the word segmentation results in the to be recognized text and the candidate intents).
Zhang does not disclose the target splitting manner is obtained based on degrees of matching between the plurality of target character sets and the plurality of first labels.
Ramaswamy discloses a conversational natural language system (Fig. 1) using a trained target model (¶¶28-29, maximum entropy model constructed from training data 70 for boundary identifier 40) to perform natural language processing of to be processed text (¶49, maximum entropy models for natural language processing) to obtain a prediction result of human intent (¶25, convert spoken command into recognized text 30 for a boundary identifier 40 to generate complete command 50 as output), the prediction result indicates to split the to be processed text into a plurality of target character sets and corresponding first labels indicating semantics of respective target character set of the target character sets (¶¶37-44, training set entries with correct decisions T=0 or T=1 (i.e., labels) indicating complete or incomplete commands; ¶¶46-49, process training data to produce feature functions 41 for Maximum Entropy Models; per ¶47, feature functions include one or more words from processed training data along with the correction decision (i.e., label); ¶51, determine which feature functions 41 are present in a given processed utterance to determine whether the given processed utterance is a complete command; per ¶26, natural language understanding system translates complete command 50 to generate formal command), and the prediction result indicates a target splitting manner corresponding to the to be processed text and the target splitting manner is obtained based on a degree of matching between the plurality of target character sets and the plurality of first labels (¶¶51-57, determine which feature functions 41 are present in a given processed utterance; e.g., ¶¶79-81, determine which feature functions 41 are present in “check for new mail show me the first one” by matching feature functions 41 with correct decision (i.e., correct labels) comprising target splitting manners such as check new mail (T = 1), check new mail show (T =0), check new mail show me (T = 1) per ¶¶54-57).
It would’ve been obvious to one ordinarily skilled in the art before the effective filing date of the invention to obtain the target splitting manner based on degrees of matching between the plurality of target character sets and the plurality of first labels (e.g., implement Ramaswamy’s feature functions including one or more words with correct decision (i.e., labels) and determine which feature functions are present in a given processed utterance per Ramaswamy, ¶51) in order to determine if a command is present in recognized text of spoken command (Ramaswamy, Abstract).
Regarding Claims 3 and 12, Zhang as modified by Ramaswamy discloses
wherein there are N splitting manners corresponding to the to-be-processed text, N is an integer greater than or equal to 1, and the target splitting manner belongs to the N splitting manners (Ramaswamy, ¶¶37-44 and ¶¶54-57, processed training data showing complete commands (T = 1) or incomplete commands (T = 0) (i.e., different manners of splitting “check new mail show me”) are used to produce feature functions per ¶46).
Regarding Claim 6, Zhang discloses a method for model training (Fig. 1), wherein the method comprises:
processing a to-be-processed text by using a target model to obtain a prediction result (¶19, acquire training text and first annotation intents of the plurality of training text; ¶20, configuring a neural network model including a first recognition layer to output a first intent result of the training text according to a first semantic vector of each segmented word in the training text outputted by a feature extraction layer of the neural network), wherein the target model is a machine learning model for natural language processing in artificial intelligence field (¶51 and Fig. 4, intent recognition model being a trained neural network comprising a first recognition layer outputs scores between word segmentation results in the to be recognized text and candidate intents and a second recognition layer processing word segmentation results to obtain second intent results), the to-be-processed text comprises a plurality of characters, the prediction result indicates to split the to-be-processed text into a plurality of target character sets, each target character set of the plurality of target character sets comprises at least one character (¶24, one example of training text is “Open the navigation app and take the highway” and word segmentation results corresponding to the training text are “open”, “navigation app”, “take”, and “highway”; ¶31, output a first intent result of the training text according to a first semantic vector of each segmented word in the training text), the prediction result further comprises a plurality of first labels, each first label of the plurality of first labels indicates semantics of a respective target character set of the plurality of target character sets (¶26, preset a plurality of candidate intents and a semantic vector corresponding to each candidate intent; ¶31, output a first intent result of the training text and a score between each segmented word in the training text and the candidate intent), the plurality of first labels are used to determine a predicted intention of the to-be-processed text (¶31, a candidate intent whose score exceeds a preset threshold is selected as the first intent result of the training text), and the prediction result indicates a target splitting manner corresponding to the to-be-processed text (¶40, train neural network model according to word segmentation results of training texts, first annotation intents, and second annotation intents; e.g., ¶24, training text “open the navigation app and take the highway” with word segmentations “open”, “navigation app”, “take”, and “highway”, first annotation intents include “NAVI” and “HIGHWAY”, and second annotation intents include “NAVI”, “NAVI”, “HIGHWAY”, and “HIGHWAY”) and
training the target model according to a target loss function to obtain a trained target model, wherein the target loss function indicates a similarity between the prediction result and an expected result corresponding to the to-be-processed text (¶35, calculating a loss function value according to the first intent results of the plurality of training texts and the first annotation intents (“expected result”) of the plurality of training texts, and using the loss function value to obtain the intent recognition model), the expected result corresponding to the to-be-processed text indicates to split the to-be-processed text into a plurality of second character sets, each second character set comprises at least one character, the expected result corresponding to the to-be-processed text further comprises a plurality of expected labels, one expected label indicates semantics of one second character set, and the plurality of expected labels are used to determine a correct intention of the to-be-processed text (¶24, in the example training text “Open the navigation app and take the highway”, word segmentation results are “open”, “navigation app”, “take” and “highway”, the annotation intent of the training text includes “NAVI” corresponding to “open” and “navigation app”, “HIGHWAY” corresponding to “take” and “highway”; i.e., second character set comprises characters “open” and “navigation app” corresponding to the annotation intent / label “NAVI”, and characters “take” and “highway” corresponding to the annotation intent / label “HIGHWAY”).
Zhang does not disclose the target splitting manner is obtained based on degrees of matching between the plurality of target character sets and the plurality of first labels.
Ramaswamy discloses a conversational natural language system (Fig. 1) training target model (¶¶28-29, maximum entropy model constructed from training data 70 for boundary identifier 40) to perform natural language processing of to be processed text (¶49, maximum entropy models for natural language processing) to obtain a prediction result of human intent (¶25, convert spoken command into recognized text 30 for a boundary identifier 40 to generate complete command 50 as output), the prediction result indicates to split the to be processed text into a plurality of target character sets and corresponding first labels indicating semantics of respective target character set of the target character sets (¶¶37-44, training set entries with correct decisions T=0 or T=1 (i.e., labels) indicating complete or incomplete commands; ¶¶46-49, process training data to produce feature functions 41 for Maximum Entropy Models; per ¶47, feature functions include one or more words from processed training data along with the correction decision (i.e., label); ¶51, determine which feature functions 41 are present in a given processed utterance to determine whether the given processed utterance is a complete command; per ¶26, natural language understanding system translates complete command 50 to generate formal command), and the prediction result indicates a target splitting manner corresponding to the to be processed text and the target splitting manner is obtained based on a degree of matching between the plurality of target character sets and the plurality of first labels (¶¶51-57, determine which feature functions 41 are present in a given processed utterance; e.g., ¶¶79-81, determine which feature functions 41 are present in “check for new mail show me the first one” by matching feature functions 41 with correct decision (i.e., correct labels) comprising target splitting manners such as check new mail (T = 1), check new mail show (T =0), check new mail show me (T = 1) per ¶¶54-57).
It would’ve been obvious to one ordinarily skilled in the art before the effective filing date of the invention to train the neural network model to obtain the target splitting manner based on degrees of matching between the plurality of target character sets and the plurality of first labels (e.g., implement Ramaswamy’s feature functions including one or more words with correct decision (i.e., labels) and determine which feature functions are present in a given processed utterance per Ramaswamy, ¶51) in order to determine if a command is present in recognized text of spoken command (Ramaswamy, Abstract) by building the model from training data (Ramaswamy, ¶29 and ¶46).
Regarding Claim 8, Zhang discloses wherein before the processing a to-be-processed text by using a target model, the method further comprises:
obtaining a target data subset, wherein the target data subset comprises a first subset and a second subset, the first subset comprises a first character string and a first expected label corresponding to the first character string, and the second subset comprises a second character string and a second expected label corresponding to the second character string (¶38, acquire training data including the plurality of training texts, the first annotation intents of the plurality of training texts and second annotation intents of the plurality of training texts); and
determining, based on the target data subset, the to-be-processed text and the expected result corresponding to the to-be-processed text, wherein the to-be-processed text comprises the first character string and the second character string, and the expected result comprises the first expected label corresponding to the first character string and the second expected label corresponding to the second character string (¶38, training texts corresponding to the first annotation intents (i.e., first expected label corresponding to first character string) and training texts corresponding to the second annotation intents (i.e., second expected label corresponding to second character string)).
Regarding Claim 9, Zhang discloses wherein a quality score corresponding to the to-be-processed text meets a preset condition, and the quality score indicates quality of the to-be-processed text (¶31, obtaining, for each training text according to a first semantic vector of each segmented word in the training text and the semantic vector of the candidate intent, a second semantic vector of each segmented word and a score between each segmented word and the candidate intent, wherein the score between each segmented word and the candidate intent may be an attention score between the two; determine candidate intent whose score exceeds a preset threshold for selection as the first intent result of the training text).
Claims 2, 7, and 11 are rejected under 35 USC 103(a) as being unpatentable over Zhang et al. (US 2023/0004798 A1) and Ramaswamy et al. (US 2014/0156268 A1) as applied to claims 1, 6, and 10, in view of Pitschel et al. (US 9922642 B2).
Regarding Claims 2, 7, and 11, Zhang does not disclose wherein the plurality of first labels comprise at least two levels of labels, the at least two levels of labels comprise a parent label and a child label, and a belonging relationship exists between the parent label and the child label.
Pitschel teaches a device for training a digital assistant (Abstract), the device uses a model process to-be-processed text to obtain a prediction result comprising a plurality of first labels (Col 10, Rows 6-15, natural language processing module 332 associates sequence of words or tokens with one or more actionable intent), wherein the plurality of first labels comprise at least two levels of labels, the at least two levels of labels comprise a parent label and a child label, and a belonging relationship exists between the parent label and the child label (Col 11, Rows 12-33, an actionable intent node along with its linked property nodes; in one example, actionable intent node (i.e., parent labels) “restaurant reservation” with property nodes (i.e., child labels) “restaurant”, “date/time”, “party size”).
It would’ve been obvious to one ordinarily skilled in the art before the effective filing date of the invention to obtain prediction result comprising a plurality of first labels, the first labels comprising at least two levels of labels comprising parent label and child label, in order to determine an intention of the to be processed text (Pitschel, Col 10, Rows 6-11).
Claims 4-5 and 13-14 are rejected under 35 USC 103(a) as being unpatentable over Zhang et al. (US 2023/0004798 A1) and Ramaswamy et al. (US 2014/0156268 A1) as applied to claims 3 and 12, in further view of Xu et al. (CN 108959257 B, see IP.com translation).
Regarding Claims 4-5 and 13-14, Zhang does not teach wherein the processing the to-be-processed text by using a target model to obtain a prediction result comprises match each target character set with a plurality of character strings in a first data set, to determine a target character string that matches the target character set.
Xu discloses a natural language parsing device for processing to-be-processed text by splitting the to-be-processed text into a plurality of target character sets and to obtain corresponding intent prediction result (Abstract, cutting words of a natural language text into word cutting segment, carry out concept labeling on each word segments to obtain concept labels, arranging and combining the concept labels into concept label sequences to carry out intention deduction) comprising:
match each target character set with a plurality of character strings in a first data set, to determine a target character string that matches the target character set (p. 9, ¶4, match each word segment with a pre-established knowledge word list; p. 9, ¶5, “the knowledge word list comprises a plurality of concept labels and phrases corresponding to the concept labels”; i.e., match each word segment with phrases in the pre-established knowledge word list);
obtain, from the first data set, at least one second label corresponding to the target character string, wherein one character string comprises at least one character (p. 9, ¶5, “one concept label in the knowledge word list is a team, and the phrases corresponding to the lower side of the team comprise the names of teams such as a Chinese team, a British team, a Germany team and the like”); and
match, based on each target character set and the second label by using the target model, each target character set with a plurality of labels in the first data set, to obtain a first label that matches each target character set (p. 9, ¶6, “And S230, if the matched concept label exists in the knowledge word list, using the matched concept label as the at least one concept label”);
wherein the at least one second label comprises at least two second labels (p. 9, ¶5, “The knowledge word list comprises a plurality of concept labels and phrases corresponding to the concept labels”), after the obtaining, from the first data set, at least one second label corresponding to the target character string, the method further comprises
generate target indication information based on the to-be-processed text, the target character set, and at least two second labels by using the target model, wherein the target indication information indicates that each second label matches or does not match the target character set (p. 9, ¶7, “if the concept label corresponding to “beijing” in the knowledge vocabulary is “city”, and “road” does not having a matching result in the knowledge vocabulary, then the word segmentation of “road” is not labeled, and if the word of “Beijing road” also exists in the knowledge vocabulary and corresponds to the concept label of “road”, then the concept label of “road” is also labeled on the combination of the adjacent word segmentation of “Beijing road”.”);
screen the at least two second labels based on the target indication information, to obtain at least one screened label (p. 9, ¶7, “Therefore, all word segmentation segments and the combination of adjacent word segmentation segments with matched concept labels in the knowledge word list can be labeled”); and
the matching, based on each target character set and the at least one second label by using the target model, each target character set with a plurality of labels in the first data set comprises match, based on the target character set and the at least one screened label by using the target model, the target character set with the plurality of labels in the first data set (p. 9, ¶8, “And S240, arranging and combining the at least one concept label to obtain a plurality of concept label sequences, wherein word cutting boundaries covered by the concept labels in each concept label sequence are not overlapped among different concept label sequences”).
It would’ve been obvious to one ordinarily skilled in the art before the effective filing date of the invention to match each target character set with a plurality of character strings in a first data set to determine a target character string that matches the target character set, obtain, from the first data set, at least one second label corresponding to the target character string, wherein one character string comprises at least one character, and match, based on each target character set and the second label by using the target model, each target character set with a plurality of labels in the first data set, to obtain a first label that matches each target character set in order to optimize natural language parsing (Xu, p. 9, ¶2, “In this embodiment, optimization is performed based on the above embodiment”).
Conclusion
Applicant's amendment necessitated the new grounds of rejection presented in this Office action. Accordingly, THIS ACTION IS MADE FINAL. See MPEP § 706.07(a). Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any extension fee pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to examiner Richard Z. Zhu whose telephone number is 571-270-1587 or examiner’s supervisor Hai Phan whose telephone number is 571-272-6338. Examiner Richard Zhu can normally be reached on M-Th, 0730:1700.
Information regarding the status of an application may be obtained from the Patent Application Information Retrieval (PAIR) system. Status information for published applications may be obtained from either Private PAIR or Public PAIR. Status information for unpublished applications is available through Private PAIR only. For more information about the PAIR system, see http://pair-direct.uspto.gov. Should you have questions on access to the Private PAIR system, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative or access to the automated information system, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/RICHARD Z ZHU/Primary Examiner, Art Unit 2654 05/22/2026
Prosecution Insights