DETAILED ACTION
This office action is in response to the application filed on February 28, 2024.
Claims 1-20 are pending and have been examined. Claims 1-20 are rejected.
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 .
Information Disclosure Statement
Acknowledgment is made of the information disclosure statement filed January 2, 2026, which complies with 37 CFR 1.97. As such, the information disclosure statement has been placed in the application file and the information referred to therein has been considered by the examiner.
Claim Objections
Claim(s) 14-15 are objected to because of the following informalities:
In Claim 14, the recitation of “a hierarchical taxonomy automatically generate based on user input” is grammatically incorrect and appears to be misspelled as “generate” instead of “generated”. It appears these recitations should read “a hierarchical taxonomy automatically generated based on user input”. Appropriate correction is required.
In independent Claim 15, the recitation of “provide the input to a first supervised machine learning (ML) classifier trained using training data to classify data into a first level of the first taxonomy” should state “a first taxonomy” instead of “the first taxonomy”. Appropriate correction is required.
Claim Rejections - 35 USC § 112
The following is a quotation of 35 U.S.C. § 112(b):
(b) CONCLUSION – The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the inventor or a joint inventor regards as the invention.
The following is a quotation of 35 U.S.C. § 112 (pre-AIA ), second paragraph:
The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the applicant regards as his invention.
Claim(s) 3 and 16-20 are rejected under 35 U.S.C. 112(b) as being indefinite for failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor (or for applications subject to pre-AIA 35 U.S.C. 112, the applicant), regards as the invention.
Regarding Claim(s) 3, there is insufficient antecedent basis for the limitations in the claim, rendering the claim indefinite because “the first output” and “the second output” are not defined in Claim 1, which Claim 3 is dependent on.
Regarding Claim(s) 16-20, there is insufficient antecedent basis for the limitations in the claims, rendering the claims indefinite because “the first taxonomy” is not defined in Claim 15, which Claims 16-20 are dependent on.
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 the claimed invention is directed to non-statutory subject matter.
According to the USPTO guidelines, a claim is directed to non-statutory subject matter if:
Step 1: The claim does not fall within one of the four statutory categories of invention (process, machine, manufacture, or composition of matter) – see MPEP 2106.03, or,
Step 2: The claim recites a judicial exception, e.g. an abstract idea, without reciting additional elements that amount to significantly more than the judicial exception, as determined using the following analysis – see MPEP 2106.04:
Step 2A, Prong 1: Does the claim recite an abstract idea, law of nature, or natural phenomenon?
Step 2A, Prong 2: Does the claim recite additional elements that integrate the judicial exception into a practical application?
Step 2B: Does the claim recite additional elements that amount to significantly more than the judicial exception? - see MPEP 2106.05
MPEP 2106.04(a)(2)(I) states: “The mathematical concepts grouping is defined as mathematical relationships, mathematical formulas or equations, and mathematical calculations.”
MPEP 2106.04(a)(2)(III) states: “Accordingly, the “mental processes” abstract idea grouping is defined as concepts performed in the human mind, and examples of mental processes include observations, evaluations, judgements, and opinions.
Further, the MPEP states: “The courts do not distinguish between mental processes that are performed entirely in the human mind and mental processes that require a human to use a physical aid (e.g. pen and paper or a slide run) to perform the claim limitation.
Using the two-step inquiry, it is clear that Claims 1-20 are each directed to non-statutory subject matter as shown below:
With respect to Claim(s) 1:
Step 1: Claim 1 is directed to a method, also known as a process, which is one of the four statutory categories of patentable subject matter.
Step 2A, Prong 1: A judicial exception is recited in the claim(s) as it recites mental processes, which are abstract ideas:
“A method for generating an artificial intelligence (AI) classifier, the method comprising: determining a first taxonomy comprising a set of nodes;” (Covers mental concepts that could be practically performed in the human mind or through the use of a pencil and paper, including observation, evaluation, and judgement - see MPEP 2106.04.)
“determining a supervised machine learning (ML) taxonomy comprising the first subset;” (Covers mental concepts that could be practically performed in the human mind or through the use of a pencil and paper, including observation, evaluation, and judgement - see MPEP 2106.04.)
“training a first supervised ML classifier based on the training data, the first supervised ML classifier trained to classify data into a first level of the supervised ML taxonomy;” (Classifying data into a first level of the supervised ML taxonomy covers mental concepts that could be practically performed in the human mind or through the use of a pencil and paper, including observation, evaluation, and judgement - see MPEP 2106.04.)
“determining a large language model (LLM) taxonomy comprising definitions for a second subset of the set of nodes;” (Covers mental concepts that could be practically performed in the human mind or through the use of a pencil and paper, including observation, evaluation, and judgement - see MPEP 2106.04.)
Step 2A, Prong 2: The claim(s) does not recite additional elements that integrate the judicial exception into a practical application:
“receiving training data for a first subset of the set of nodes;” (Receiving training data for a first subset of the set of nodes is akin to obtaining data, which adds insignificant extra-solution activity to the judicial exception, as discussed in MPEP 2106.05(g).)
“determining a supervised machine learning (ML) taxonomy comprising the first subset;” (Supervised machine learning (ML) only amounts to “apply it” and mere instructions to implement an abstract idea on a computer – see MPEP 2106.05(f)(1).)
“training a first supervised ML classifier based on the training data, the first supervised ML classifier trained to classify data into a first level of the supervised ML taxonomy;” (Training a first supervised ML classifier based on the training data only amount to “apply it” and mere instructions to implement an abstract idea on a computer – see MPEP 2106.05(f)(1).)
“determining a large language model (LLM) taxonomy comprising definitions for a second subset of the set of nodes;” (Large language models (LLM) only amount to “apply it” and mere instructions to implement an abstract idea on a computer – see MPEP 2106.05(f)(1).)
“and deploying a hybrid classifier enabled to classify an input based on a first classification determined, by the first supervised ML classifier, based on the input, and a second classification generated, by the LLM, based on a prompt comprising the input and at least a portion of the LLM taxonomy.” (Deploying a hybrid classifier enabled to classify an input based on a first classification determined, by the first supervised ML classifier, based on the input, and a second classification generated, by the LLM, based on a prompt comprising the input and at least a portion of the LLM taxonomy only amount to “apply it” and mere instructions to implement an abstract idea on a computer – see MPEP 2106.05(f)(1).)
Step 2B: The claim(s) does not recite additional elements that amount to significantly more than the judicial exception. Receiving training data for a first subset of the set of nodes is akin to obtaining data, which adds insignificant extra-solution activity to the judicial exception, as discussed in MPEP 2106.05(g). Furthermore, the additional element is directed to receiving or transmitting data over a network / performing repetitive calculations / electronic recordkeeping / storing and retrieving information in memory / electronically scanning or extracting data from a physical document, which the courts have recognized as well‐understood, routine, and conventional when they are claimed in a generic manner. See MPEP § 2106.05(d)(II).
Supervised machine learning (ML) only amounts to “apply it” and mere instructions to implement an abstract idea on a computer – see MPEP 2106.05(f)(1). Training a first supervised ML classifier based on the training data only amount to “apply it” and mere instructions to implement an abstract idea on a computer – see MPEP 2106.05(f)(1). Large language models (LLM) only amount to “apply it” and mere instructions to implement an abstract idea on a computer – see MPEP 2106.05(f)(1). Deploying a hybrid classifier enabled to classify an input based on a first classification determined, by the first supervised ML classifier, based on the input, and a second classification generated, by the LLM, based on a prompt comprising the input and at least a portion of the LLM taxonomy only amount to “apply it” and mere instructions to implement an abstract idea on a computer – see MPEP 2106.05(f)(1).
With respect to Claim(s) 2:
Step 2A, Prong 1: Inherits the limitations and abstract ideas from Claim 1. An additional judicial exception is recited in the claim(s) as it recites mental processes, which are abstract ideas:
“and determining a first level classification of the input in the first level of the first taxonomy based on the second classification.” (Covers mental concepts that could be practically performed in the human mind or through the use of a pencil and paper, including observation, evaluation, and judgement - see MPEP 2106.04.)
Step 2A, Prong 2: The claim(s) does not recite additional elements that integrate the judicial exception into a practical application:
“wherein the hybrid classifier classifies the input by: providing the input to the first supervised ML classifier;” (Providing the input to the first supervised ML classifier is akin to inputting data, which adds insignificant extra-solution activity to the judicial exception, as discussed in MPEP 2106.05(g).)
“receiving, from the first supervised ML classifier, the first classification;” (Receiving, from the first supervised ML classifier, the first classification is akin to receiving data, which adds insignificant extra-solution activity to the judicial exception, as discussed in MPEP 2106.05(g).)
“providing, to the LLM, the prompt comprising the input, a first portion of the LLM taxonomy, and the first classification;” (Providing, to the LLM, the prompt comprising the input, a first portion of the LLM taxonomy, and the first classification is akin to inputting data, which adds insignificant extra-solution activity to the judicial exception, as discussed in MPEP 2106.05(g).)
“receiving, from the LLM, the second classification;” (Receiving, from the LLM, the second classification is akin to outputting data, which adds insignificant extra-solution activity to the judicial exception, as discussed in MPEP 2106.05(g).)
Step 2B: The claim(s) does not recite additional elements that amount to significantly more than the judicial exception. Providing the input to the first supervised ML classifier is akin to inputting data, which adds insignificant extra-solution activity to the judicial exception, as discussed in MPEP 2106.05(g). Furthermore, the additional element is directed to receiving or transmitting data over a network / performing repetitive calculations / electronic recordkeeping / storing and retrieving information in memory / electronically scanning or extracting data from a physical document, which the courts have recognized as well‐understood, routine, and conventional when they are claimed in a generic manner. See MPEP § 2106.05(d)(II).
Receiving, from the first supervised ML classifier, the first classification is akin to receiving data, which adds insignificant extra-solution activity to the judicial exception, as discussed in MPEP 2106.05(g). Providing, to the LLM, the prompt comprising the input, a first portion of the LLM taxonomy, and the first classification is akin to inputting data, which adds insignificant extra-solution activity to the judicial exception, as discussed in MPEP 2106.05(g). Furthermore, the additional element is directed to receiving or transmitting data over a network / performing repetitive calculations / electronic recordkeeping / storing and retrieving information in memory / electronically scanning or extracting data from a physical document, which the courts have recognized as well‐understood, routine, and conventional when they are claimed in a generic manner. See MPEP § 2106.05(d)(II).
Receiving, from the LLM, the second classification is akin to outputting data, which adds insignificant extra-solution activity to the judicial exception, as discussed in MPEP 2106.05(g). Furthermore, the additional element is directed to receiving or transmitting data over a network / performing repetitive calculations / electronic recordkeeping / storing and retrieving information in memory / electronically scanning or extracting data from a physical document, which the courts have recognized as well‐understood, routine, and conventional when they are claimed in a generic manner. See MPEP § 2106.05(d)(II).
With respect to Claim(s) 3:
Step 2A, Prong 1: Inherits the limitations and abstract ideas from Claim 1. An additional judicial exception is recited in the claim(s) as it recites mental processes, which are abstract ideas:
“and determining a first level classification of the input in the first level of the first taxonomy based on the first output, the second output, the first confidence score, and the second confidence score.” (Covers mental concepts that could be practically performed in the human mind or through the use of a pencil and paper, including observation, evaluation, and judgement - see MPEP 2106.04.)
Step 2A, Prong 2: The claim does not recite additional elements that integrate the judicial exception into a practical application:
“wherein the hybrid classifier classifies the input by: providing the input to the first supervised ML classifier;” (Providing the input to the first supervised ML classifier is akin to inputting data, which adds insignificant extra-solution activity to the judicial exception, as discussed in MPEP 2106.05(g).)
“receiving, from the first supervised ML classifier, the first classification and a first confidence score associated with the first classification;” (Receiving, from the first supervised ML classifier, the first classification and a first confidence score associated with the first classification is akin to receiving data, which adds insignificant extra-solution activity to the judicial exception, as discussed in MPEP 2106.05(g).)
“providing, to the LLM, the prompt comprising the input and a first portion of the LLM taxonomy;” (Providing, to the LLM, the prompt comprising the input and a first portion of the LLM taxonomy is akin to inputting data, which adds insignificant extra-solution activity to the judicial exception, as discussed in MPEP 2106.05(g).)
“receiving, from the LLM, the second classification and a second confidence score associated with the second classification;” (Receiving, from the LLM, the second classification and a second confidence score associated with the second classification is akin to receiving data, which adds insignificant extra-solution activity to the judicial exception, as discussed in MPEP 2106.05(g).)
Step 2B: The claim does not recite additional elements that amount to significantly more than the judicial exception. Providing the input to the first supervised ML classifier is akin to inputting data, which adds insignificant extra-solution activity to the judicial exception, as discussed in MPEP 2106.05(g). Furthermore, the additional element is directed to receiving or transmitting data over a network / performing repetitive calculations / electronic recordkeeping / storing and retrieving information in memory / electronically scanning or extracting data from a physical document, which the courts have recognized as well‐understood, routine, and conventional when they are claimed in a generic manner. See MPEP § 2106.05(d)(II).
Receiving, from the first supervised ML classifier, the first classification and a first confidence score associated with the first classification is akin to receiving data, which adds insignificant extra-solution activity to the judicial exception, as discussed in MPEP 2106.05(g). Furthermore, the additional element is directed to receiving or transmitting data over a network / performing repetitive calculations / electronic recordkeeping / storing and retrieving information in memory / electronically scanning or extracting data from a physical document, which the courts have recognized as well‐understood, routine, and conventional when they are claimed in a generic manner. See MPEP § 2106.05(d)(II).
Providing, to the LLM, the prompt comprising the input and a first portion of the LLM taxonomy is akin to inputting data, which adds insignificant extra-solution activity to the judicial exception, as discussed in MPEP 2106.05(g). Furthermore, the additional element is directed to receiving or transmitting data over a network / performing repetitive calculations / electronic recordkeeping / storing and retrieving information in memory / electronically scanning or extracting data from a physical document, which the courts have recognized as well‐understood, routine, and conventional when they are claimed in a generic manner. See MPEP § 2106.05(d)(II).
Receiving, from the LLM, the second classification and a second confidence score associated with the second classification is akin to receiving data, which adds insignificant extra-solution activity to the judicial exception, as discussed in MPEP 2106.05(g). Furthermore, the additional element is directed to receiving or transmitting data over a network / performing repetitive calculations / electronic recordkeeping / storing and retrieving information in memory / electronically scanning or extracting data from a physical document, which the courts have recognized as well‐understood, routine, and conventional when they are claimed in a generic manner. See MPEP § 2106.05(d)(II).
With respect to Claim(s) 4:
Step 2A, Prong 1: Inherits the limitations and abstract ideas from Claim 1. An additional judicial exception is recited in the claim(s) as it recites mental processes, which are abstract ideas:
“determining a second level classification of the input in the second level of the first taxonomy based on the third classification and the fourth classification, the second level of the first taxonomy being subordinate to the first level of the first taxonomy;” (Covers mental concepts that could be practically performed in the human mind or through the use of a pencil and paper, including observation, evaluation, and judgement - see MPEP 2106.04.)
Step 2A, Prong 2: The claim does not recite additional elements that integrate the judicial exception into a practical application:
“training a second supervised ML classifier based on the training data, the second supervised ML classifier trained to classify data into a second level of the supervised ML taxonomy;” (Training a second supervised ML classifier based on the training data, the second supervised ML classifier trained to classify data into a second level of the supervised ML taxonomy only amount to “apply it” and mere instructions to implement an abstract idea on a computer – see MPEP 2106.05(f)(1).)
“providing, to the second supervised ML classifier, the input and the first level classification; receiving, from the second supervised ML classifier, a third classification of the input;” (Providing, to the second supervised ML classifier, the input and the first level classification; receiving, from the second supervised ML classifier, a third classification of the input is akin to inputting data, which adds insignificant extra-solution activity to the judicial exception, as discussed in MPEP 2106.05(g).)
“providing, to a large language model (LLM), a second prompt comprising the input, the first level classification, and a second portion of the LLM taxonomy;” (Providing, to a large language model (LLM), a second prompt comprising the input, the first level classification, and a second portion of the LLM taxonomy is akin to inputting data, which adds insignificant extra-solution activity to the judicial exception, as discussed in MPEP 2106.05(g).)
“receiving, from the LLM, a fourth classification determined by the LLM based at least on the input, the first level classification, and the second portion of the LLM taxonomy;” (Receiving, from the LLM, a fourth classification determined by the LLM based at least on the input, the first level classification, and the second portion of the LLM taxonomy is akin to receiving data, which adds insignificant extra-solution activity to the judicial exception, as discussed in MPEP 2106.05(g).)
“and providing, as an overall classification of the input, an output classification comprising the first level classification and the second level classification.” (Providing, as an overall classification of the input, an output classification comprising the first level classification and the second level classification is akin to inputting data, which adds insignificant extra-solution activity to the judicial exception, as discussed in MPEP 2106.05(g).)
Step 2B: The claim does not recite additional elements that amount to significantly more than the judicial exception. Training a second supervised ML classifier based on the training data, the second supervised ML classifier trained to classify data into a second level of the supervised ML taxonomy only amount to “apply it” and mere instructions to implement an abstract idea on a computer – see MPEP 2106.05(f)(1).
Providing, to the second supervised ML classifier, the input and the first level classification; receiving, from the second supervised ML classifier, a third classification of the input is akin to inputting data, which adds insignificant extra-solution activity to the judicial exception, as discussed in MPEP 2106.05(g). Furthermore, the additional element is directed to receiving or transmitting data over a network / performing repetitive calculations / electronic recordkeeping / storing and retrieving information in memory / electronically scanning or extracting data from a physical document, which the courts have recognized as well‐understood, routine, and conventional when they are claimed in a generic manner. See MPEP § 2106.05(d)(II).
Providing, to a large language model (LLM), a second prompt comprising the input, the first level classification, and a second portion of the LLM taxonomy is akin to inputting data, which adds insignificant extra-solution activity to the judicial exception, as discussed in MPEP 2106.05(g). Furthermore, the additional element is directed to receiving or transmitting data over a network / performing repetitive calculations / electronic recordkeeping / storing and retrieving information in memory / electronically scanning or extracting data from a physical document, which the courts have recognized as well‐understood, routine, and conventional when they are claimed in a generic manner. See MPEP § 2106.05(d)(II).
Receiving, from the LLM, a fourth classification determined by the LLM based at least on the input, the first level classification, and the second portion of the LLM taxonomy is akin to receiving data, which adds insignificant extra-solution activity to the judicial exception, as discussed in MPEP 2106.05(g). Furthermore, the additional element is directed to receiving or transmitting data over a network / performing repetitive calculations / electronic recordkeeping / storing and retrieving information in memory / electronically scanning or extracting data from a physical document, which the courts have recognized as well‐understood, routine, and conventional when they are claimed in a generic manner. See MPEP § 2106.05(d)(II).
Providing, as an overall classification of the input, an output classification comprising the first level classification and the second level classification is akin to inputting data, which adds insignificant extra-solution activity to the judicial exception, as discussed in MPEP 2106.05(g). Furthermore, the additional element is directed to receiving or transmitting data over a network / performing repetitive calculations / electronic recordkeeping / storing and retrieving information in memory / electronically scanning or extracting data from a physical document, which the courts have recognized as well‐understood, routine, and conventional when they are claimed in a generic manner. See MPEP § 2106.05(d)(II).
With respect to Claim(s) 5:
Step 2A, Prong 1: Inherits the limitations and abstract ideas from Claim 1. An additional judicial exception is recited in the claim(s) as it recites mental processes, which are abstract ideas:
“determining a definition for a node of the second subset based on an analysis of the training data associated with the first subset;” (Covers mental concepts that could be practically performed in the human mind or through the use of a pencil and paper, including observation, evaluation, and judgement - see MPEP 2106.04.)
“or determining a definition for a node of the second subset based on information associated with ancestor nodes of the node of the second subset.” (Covers mental concepts that could be practically performed in the human mind or through the use of a pencil and paper, including observation, evaluation, and judgement - see MPEP 2106.04.)
Step 2A, Prong 2: The claim does not recite additional elements that integrate the judicial exception into a practical application:
“wherein determining the LLM taxonomy comprises at least one of: receiving, from a user via a user interface, a definition for a node of the second subset;” (Receiving a definition for a node of a second subset from a user via a user interface is also akin to receiving data, which adds insignificant extra-solution activity to the judicial exception, as discussed in MPEP 2106.05(g).)
Step 2B: The claim does not recite additional elements that amount to significantly more than the judicial exception. Receiving a definition for a node of a second subset from a user via a user interface is also akin to receiving data, which adds insignificant extra-solution activity to the judicial exception, as discussed in MPEP 2106.05(g). Furthermore, the additional element is directed to receiving or transmitting data over a network / performing repetitive calculations / electronic recordkeeping / storing and retrieving information in memory / electronically scanning or extracting data from a physical document, which the courts have recognized as well‐understood, routine, and conventional when they are claimed in a generic manner. See MPEP § 2106.05(d)(II).
With respect to Claim(s) 6, 13, and 20:
Step 2A, Prong 1: Inherits the limitations and abstract ideas from Claim 1.
Step 2A, Prong 2: The claim does not recite additional elements that integrate the judicial exception into a practical application:
“wherein the training data comprises at least one of: user-provided training data associated with a node of the first taxonomy; or training data retrieved based on a user-provided document identifier and associated with a node of the first taxonomy, the user-provided document identifier comprising at least one of: a publication identifier, or a patent number.” (User-provided training data associated with a node of the first taxonomy is akin to insignificant application, which adds insignificant extra-solution activity to the judicial exception, as discussed in MPEP 2106.05(g).)
Step 2B: The claim does not recite additional elements that amount to significantly more than the judicial exception. User-provided training data associated with a node of the first taxonomy is akin to insignificant application, which adds insignificant extra-solution activity to the judicial exception, as discussed in MPEP 2106.05(g).
With respect to Claim(s) 7:
Step 2A, Prong 1: Inherits the limitations and abstract ideas from Claim 1. An additional judicial exception is recited in the claim(s) as it recites mental processes, which are abstract ideas:
“automatically generating at least a portion of the first taxonomy based on user input;” (Covers mental concepts that could be practically performed in the human mind or through the use of a pencil and paper, including observation, evaluation, and judgement - see MPEP 2106.04.)
“or generating the first taxonomy based on user modifications to a previously generated taxonomy.” (Covers mental concepts that could be practically performed in the human mind or through the use of a pencil and paper, including observation, evaluation, and judgement - see MPEP 2106.04.)
Step 2A, Prong 2: The claim does not recite additional elements that integrate the judicial exception into a practical application:
“wherein the first taxonomy is a hierarchical taxonomy, (A first taxonomy that is a hierarchical taxonomy is akin to insignificant application, which adds insignificant extra-solution activity to the judicial exception, as discussed in MPEP 2106.05(g).)
and determining the first taxonomy comprises at least one of: receiving, from a user via a user interface, at least a portion of the first taxonomy;” (Receiving at least a portion of the first taxonomy from a user via a user interface is also akin to receiving data, which adds insignificant extra-solution activity to the judicial exception, as discussed in MPEP 2106.05(g).)
Step 2B: The claim does not recite additional elements that amount to significantly more than the judicial exception. A first taxonomy that is a hierarchical taxonomy is akin to insignificant application, which adds insignificant extra-solution activity to the judicial exception, as discussed in MPEP 2106.05(g).
Receiving at least a portion of the first taxonomy from a user via a user interface is also akin to receiving data, which adds insignificant extra-solution activity to the judicial exception, as discussed in MPEP 2106.05(g). Furthermore, the additional element is directed to receiving or transmitting data over a network / performing repetitive calculations / electronic recordkeeping / storing and retrieving information in memory / electronically scanning or extracting data from a physical document, which the courts have recognized as well‐understood, routine, and conventional when they are claimed in a generic manner. See MPEP § 2106.05(d)(II).
With respect to Claim(s) 8 and 15:
Step 1: Claims 8 and 15 are directed to an apparatus, which is one of the four statutory categories of patentable subject matter.
Step 2A, Prong 1: A judicial exception is recited in the claim(s) as it recites mental processes, which are abstract ideas:
“provide the input to a first supervised machine learning (ML) classifier trained using training data to classify data into a first level of the first taxonomy;” (Covers mental concepts that could be practically performed in the human mind or through the use of a pencil and paper, including observation, evaluation, and judgement - see MPEP 2106.04.)
“and determine a first level classification of the input in the first level of the first taxonomy based on the first classification and the second classification.” (Covers mental concepts that could be practically performed in the human mind or through the use of a pencil and paper, including observation, evaluation, and judgement - see MPEP 2106.04.)
Step 2A, Prong 2: The claim(s) do not recite additional elements that integrate the judicial exception into a practical application:
“A system for classifying an input into a node of a taxonomy, the system comprising:
a processor; and a memory device that stores program code structured to cause the processor to: (A processer and a memory device that stores program code that causes a processor to perform an action is an additional element that amounts to adding the words “apply it” (or an equivalent) with the judicial exception, or merely uses a computer in its ordinary capacity as a tool to perform an existing process - see MPEP §§ 2106.04(d) and MPEP §§ 2106.05(f)(2).)
receive the input;” (Receiving the input is akin to inputting data, which adds insignificant extra-solution activity to the judicial exception, as discussed in MPEP 2106.05(g).)
“provide the input to a first supervised machine learning (ML) classifier trained using training data to classify data into a first level of the first taxonomy;” (Providing the input to a first supervised machine learning (ML) classifier trained using training data is akin to inputting data, which adds insignificant extra-solution activity to the judicial exception, as discussed in MPEP 2106.05(g).)
“receive, from the first supervised ML classifier, a first classification of the input;” (Receiving, from the first supervised ML classifier, a first classification of the input is akin to receiving data, which adds insignificant extra-solution activity to the judicial exception, as discussed in MPEP 2106.05(g).)
“provide, to a large language model (LLM), a first prompt comprising the input and a first portion of an LLM taxonomy comprising definitions for nodes of the first taxonomy;” (Providing, to a large language model (LLM), a first prompt comprising the input and a first portion of an LLM taxonomy comprising definitions for nodes of the first taxonomy is akin to inputting data, which adds insignificant extra-solution activity to the judicial exception, as discussed in MPEP 2106.05(g).)
“receive, from the LLM, a second classification of the input determined by the LLM based at least on the input and the first portion of the LLM taxonomy;” (Receiving, from the LLM, a second classification of the input determined by the LLM based at least on the input and the first portion of the LLM taxonomy is akin to receiving data, which adds insignificant extra-solution activity to the judicial exception, as discussed in MPEP 2106.05(g).)
Step 2B: The claim(s) does not recite additional elements that amount to significantly more than the judicial exception. A processer and a memory device that stores program code that causes a processor to perform an action is an additional element that amounts to adding the words “apply it” (or an equivalent) with the judicial exception, or merely uses a computer in its ordinary capacity as a tool to perform an existing process - see MPEP §§ 2106.04(d) and MPEP §§ 2106.05(f)(2).
Receiving the input is akin to inputting data, which adds insignificant extra-solution activity to the judicial exception, as discussed in MPEP 2106.05(g). Providing the input to a first supervised machine learning (ML) classifier trained using training data is akin to inputting data, which adds insignificant extra-solution activity to the judicial exception, as discussed in MPEP 2106.05(g). Furthermore, the additional element is directed to receiving or transmitting data over a network / performing repetitive calculations / electronic recordkeeping / storing and retrieving information in memory / electronically scanning or extracting data from a physical document, which the courts have recognized as well‐understood, routine, and conventional when they are claimed in a generic manner. See MPEP § 2106.05(d)(II).
Receiving, from the first supervised ML classifier, a first classification of the input is akin to receiving data, which adds insignificant extra-solution activity to the judicial exception, as discussed in MPEP 2106.05(g). Providing, to a large language model (LLM), a first prompt comprising the input and a first portion of an LLM taxonomy comprising definitions for nodes of the first taxonomy is akin to inputting data, which adds insignificant extra-solution activity to the judicial exception, as discussed in MPEP 2106.05(g). Furthermore, the additional element is directed to receiving or transmitting data over a network / performing repetitive calculations / electronic recordkeeping / storing and retrieving information in memory / electronically scanning or extracting data from a physical document, which the courts have recognized as well‐understood, routine, and conventional when they are claimed in a generic manner. See MPEP § 2106.05(d)(II).
Receiving, from the LLM, a second classification of the input determined by the LLM based at least on the input and the first portion of the LLM taxonomy is akin to receiving data, which adds insignificant extra-solution activity to the judicial exception, as discussed in MPEP 2106.05(g). Furthermore, the additional element is directed to receiving or transmitting data over a network / performing repetitive calculations / electronic recordkeeping / storing and retrieving information in memory / electronically scanning or extracting data from a physical document, which the courts have recognized as well‐understood, routine, and conventional when they are claimed in a generic manner. See MPEP § 2106.05(d)(II).
With respect to Claim(s) 9 and 16:
Step 2A, Prong 1: Inherits the limitations and abstract ideas from Claim(s) 8 and 15, respectively. An additional judicial exception is recited in the claim(s) as it recites mental processes, which are abstract ideas:
“and determine the second classification as the first level classification of the input in the first level of the first taxonomy.” (Covers mental concepts that could be practically performed in the human mind or through the use of a pencil and paper, including observation, evaluation, and judgement - see MPEP 2106.04.)
Step 2A, Prong 2: The claim does not recite additional elements that integrate the judicial exception into a practical application:
“wherein, to determine a first level classification of the input in the first level of the first taxonomy based on the first classification and the second classification, the program code is structured to further cause the processor to: provide the first classification in the prompt to the LLM, the LLM determining the second classification further based on the first classification;” (Providing the first classification in the prompt to the LLM, the LLM determining the second classification further based on the first classification is akin to inputting data, which adds insignificant extra-solution activity to the judicial exception, as discussed in MPEP 2106.05(g).)
Step 2B: The claim does not recite additional elements that amount to significantly more than the judicial exception. Providing the first classification in the prompt to the LLM, the LLM determining the second classification further based on the first classification is akin to inputting data, which adds insignificant extra-solution activity to the judicial exception, as discussed in MPEP 2106.05(g). Furthermore, the additional element is directed to receiving or transmitting data over a network / performing repetitive calculations / electronic recordkeeping / storing and retrieving information in memory / electronically scanning or extracting data from a physical document, which the courts have recognized as well‐understood, routine, and conventional when they are claimed in a generic manner. See MPEP § 2106.05(d)(II).
With respect to Claim(s) 10 and 17:
Step 2A, Prong 1: Inherits the limitations and abstract ideas from Claim(s) 8 and 15, respectively. An additional judicial exception is recited in the claim(s) as it recites mental processes, which are abstract ideas:
“and determine the first level classification of the input in the first level of the first taxonomy based on the first classification, the second classification, the first confidence score, and the second confidence score.” (Covers mental concepts that could be practically performed in the human mind or through the use of a pencil and paper, including observation, evaluation, and judgement - see MPEP 2106.04.)
Step 2A, Prong 2: The claim does not recite additional elements that integrate the judicial exception into a practical application:
“wherein, to determine a first level classification of the input in the first level of the first taxonomy based on the first classification and the second classification, the program code is structured to further cause the processor to: receive, from the first supervised ML classifier, a first confidence score associated with the first classification;” (Receiving, from the first supervised ML classifier, a first confidence score associated with the first classification is akin to receiving data, which adds insignificant extra-solution activity to the judicial exception, as discussed in MPEP 2106.05(g).)
“receive, from the LLM, a second confidence score associated with the second classification;” (Receiving, from the LLM, a second confidence score associated with the second classification is akin to outputting data, which adds insignificant extra-solution activity to the judicial exception, as discussed in MPEP 2106.05(g).)
Step 2B: The claim does not recite additional elements that amount to significantly more than the judicial exception. Receiving, from the first supervised ML classifier, a first confidence score associated with the first classification is akin to receiving data, which adds insignificant extra-solution activity to the judicial exception, as discussed in MPEP 2106.05(g). Furthermore, the additional element is directed to receiving or transmitting data over a network / performing repetitive calculations / electronic recordkeeping / storing and retrieving information in memory / electronically scanning or extracting data from a physical document, which the courts have recognized as well‐understood, routine, and conventional when they are claimed in a generic manner. See MPEP § 2106.05(d)(II).
Receiving, from the LLM, a second confidence score associated with the second classification is akin to outputting data, which adds insignificant extra-solution activity to the judicial exception, as discussed in MPEP 2106.05(g). Furthermore, the additional element is directed to receiving or transmitting data over a network / performing repetitive calculations / electronic recordkeeping / storing and retrieving information in memory / electronically scanning or extracting data from a physical document, which the courts have recognized as well‐understood, routine, and conventional when they are claimed in a generic manner. See MPEP § 2106.05(d)(II).
With respect to Claim(s) 11 and 18:
Step 2A, Prong 1: Inherits the limitations and abstract ideas from Claim(s) 8 and 15, respectively. An additional judicial exception is recited in the claim(s) as it recites mental processes, which are abstract ideas:
“determine a second level classification of the input in a second level of the first taxonomy based on the third classification and the fourth classification, the second level of the first taxonomy being subordinate to the first level of the first taxonomy;” (Covers mental concepts that could be practically performed in the human mind or through the use of a pencil and paper, including observation, evaluation, and judgement - see MPEP 2106.04.)
Step 2A, Prong 2: The claim does not recite additional elements that integrate the judicial exception into a practical application:
“wherein the program code is structured to further cause the processor to: provide the input and the first level classification to a second supervised ML classifier trained using the training data to classify data into a second level of the first taxonomy that is subordinate to the first level;” (Providing the input and the first level classification to a second supervised ML classifier trained using the training data to classify data into a second level of the first taxonomy that is subordinate to the first level is akin to inputting data, which adds insignificant extra-solution activity to the judicial exception, as discussed in MPEP 2106.05(g).)
“receive, from the second supervised ML classifier, a third classification for the input; (Receiving, from the second supervised ML classifier, a third classification for the input is akin to outputting data, which adds insignificant extra-solution activity to the judicial exception, as discussed in MPEP 2106.05(g).)
provide, to a large language model (LLM), a second prompt comprising the input, the first level classification, and a second portion of the LLM taxonomy;” (Providing, to a large language model (LLM), a second prompt comprising the input, the first level classification, and a second portion of the LLM taxonomy is akin to inputting data, which adds insignificant extra-solution activity to the judicial exception, as discussed in MPEP 2106.05(g).)
“receive, from the LLM, a fourth classification, the fourth classification determined by the LLM based at least on the input, the first level classification, and the second portion of the LLM taxonomy;” (Receiving, from the LLM, a fourth classification, the fourth classification determined by the LLM based at least on the input, the first level classification, and the second portion of the LLM taxonomy is akin to receiving data, which adds insignificant extra-solution activity to the judicial exception, as discussed in MPEP 2106.05(g).)
“and provide, as an overall classification of the input, an output classification comprising the first level classification and the second level classification.” (Providing, as an overall classification of the input, an output classification comprising the first level classification and the second level classification is akin to outputting data, which adds insignificant extra-solution activity to the judicial exception, as discussed in MPEP 2106.05(g).)
Step 2B: The claim does not recite additional elements that amount to significantly more than the judicial exception. Providing the input and the first level classification to a second supervised ML classifier trained using the training data to classify data into a second level of the first taxonomy that is subordinate to the first level is akin to inputting data, which adds insignificant extra-solution activity to the judicial exception, as discussed in MPEP 2106.05(g). Furthermore, the additional element is directed to receiving or transmitting data over a network / performing repetitive calculations / electronic recordkeeping / storing and retrieving information in memory / electronically scanning or extracting data from a physical document, which the courts have recognized as well‐understood, routine, and conventional when they are claimed in a generic manner. See MPEP § 2106.05(d)(II).
Receiving, from the second supervised ML classifier, a third classification for the input is akin to outputting data, which adds insignificant extra-solution activity to the judicial exception, as discussed in MPEP 2106.05(g). Furthermore, the additional element is directed to receiving or transmitting data over a network / performing repetitive calculations / electronic recordkeeping / storing and retrieving information in memory / electronically scanning or extracting data from a physical document, which the courts have recognized as well‐understood, routine, and conventional when they are claimed in a generic manner. See MPEP § 2106.05(d)(II).
Providing, to a large language model (LLM), a second prompt comprising the input, the first level classification, and a second portion of the LLM taxonomy is akin to inputting data, which adds insignificant extra-solution activity to the judicial exception, as discussed in MPEP 2106.05(g). Furthermore, the additional element is directed to receiving or transmitting data over a network / performing repetitive calculations / electronic recordkeeping / storing and retrieving information in memory / electronically scanning or extracting data from a physical document, which the courts have recognized as well‐understood, routine, and conventional when they are claimed in a generic manner. See MPEP § 2106.05(d)(II).
Receiving, from the LLM, a fourth classification, the fourth classification determined by the LLM based at least on the input, the first level classification, and the second portion of the LLM taxonomy is akin to receiving data, which adds insignificant extra-solution activity to the judicial exception, as discussed in MPEP 2106.05(g). Furthermore, the additional element is directed to receiving or transmitting data over a network / performing repetitive calculations / electronic recordkeeping / storing and retrieving information in memory / electronically scanning or extracting data from a physical document, which the courts have recognized as well‐understood, routine, and conventional when they are claimed in a generic manner. See MPEP § 2106.05(d)(II).
Providing, as an overall classification of the input, an output classification comprising the first level classification and the second level classification is akin to outputting data, which adds insignificant extra-solution activity to the judicial exception, as discussed in MPEP 2106.05(g). Furthermore, the additional element is directed to receiving or transmitting data over a network / performing repetitive calculations / electronic recordkeeping / storing and retrieving information in memory / electronically scanning or extracting data from a physical document, which the courts have recognized as well‐understood, routine, and conventional when they are claimed in a generic manner. See MPEP § 2106.05(d)(II).
With respect to Claim(s) 12 and 19:
Step 2A, Prong 1: Inherits the limitations and abstract ideas from Claim(s) 8 and 15, respectively. An additional judicial exception is recited in the claim(s) as it recites mental processes, which are abstract ideas:
“wherein the LLM taxonomy comprises at least one of: a definition for a node of the first taxonomy received from a user via a user interface; a definition for a node of the first taxonomy determined based on an analysis of training data; a definition for a node of the first taxonomy determined based on information associated with subordinate nodes of the node; or a definition for a node of the first taxonomy determined based on information associated with ancestor nodes of the node.” (Determining a definition for a node of the first taxonomy based on an analysis of training data covers mental concepts that could be practically performed in the human mind or through the use of a pencil and paper, including observation, evaluation, and judgement - see MPEP 2106.04.)
Step 2A, Prong 2: The claim does not recite additional elements that integrate the judicial exception into a practical application.
Step 2B: The claim does not recite additional elements that amount to significantly more than the judicial exception.
With respect to Claim(s) 14:
Step 2A, Prong 1: Inherits the limitations and abstract ideas from Claim(s) 8.
“wherein the first taxonomy comprises at least one of: a hierarchical taxonomy received from a user via a user interface; a hierarchical taxonomy automatically generate based on user input; or a hierarchical taxonomy generated based on user modifications to a previously generated taxonomy.” (Automatically generating a hierarchical taxonomy based on user input covers mental concepts that could be practically performed in the human mind or through the use of a pencil and paper, including observation, evaluation, and judgement - see MPEP 2106.04.)
Step 2A, Prong 2: The claim(s) does not recite additional elements that integrate the judicial exception into a practical application.
Step 2B: The claim(s) does not recite additional elements that amount to significantly more than the judicial exception.
Claim Rejections - 35 USC § 103
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis (i.e. changing from AIA to pre-AIA ) for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
This application currently names joint inventors. In considering patentability of the claims the examiner presumes that the subject matter of the various claims was commonly owned as of the effective filing date of the claimed invention(s) absent any evidence to the contrary. Applicant is advised of the obligation under 37 CFR 1.56 to point out the inventor and effective filing dates of each claim that was not commonly owned as of the effective filing date of the later invention in order for the examiner to consider the applicability of 35 U.S.C. 102(b)(2)(C) for any potential 35 U.S.C. 102(a)(2) prior art against the later invention.
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.
The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or non-obviousness.
Claim(s) 1, 3, 5-6, 8, 10, 11-13, 15, and 17-20 are rejected under 35 U.S.C. 103 as being unpatentable over Shacham et al., (US Patent Application Number US10726355B2 filed on May 31, 2016, hereinafter “Shacham”), in view of “Blueprinting the Future: Automatic Item Categorization Using Hierarchical Zero-Shot and Few-Shot Classifiers” by Wang et al., (non-patent literature published on December 6, 2023, hereinafter “Wang”).
With respect to Claim(s) 1:
Shacham teaches:
“A method for generating an artificial intelligence (AI) classifier, the method comprising:
determining a first taxonomy comprising a set of nodes;” (Column 10, Lines 7-11 recite an industry taxonomy (a first taxonomy) consisting of company industry categories and subcategories (nodes).)
“receiving training data for a first subset of the set of nodes;” (Column 6, Lines 46-51 recite utilizing a first set of features extracted from training data, where the training data consists of labels indicting appropriate industries for companies (a first subset of the set of nodes).)
“determining a supervised machine learning (ML) taxonomy comprising the first subset;” (Column 6, Lines 20-32 recite the industry taxonomy gets divided into two subsets, where the industry taxonomy is split according to two different time periods, existing industries (first subset) and new industries. Column 6, Lines 40-42 further recite that the existing industry classifier (first subset) is trained via a supervised machine learning algorithm.)
“training a first supervised ML classifier based on the training data, the first supervised ML classifier trained to classify data into a first level of the supervised ML taxonomy;” (Column 6, Lines 40-42 recite that the existing industry classifier is trained via a supervised machine learning algorithm. (training a first supervised ML classifier based on the training data). Column 6, Lines 50-52 recite the labeled training data fed into the first machine learning algorithm (supervised) consists of a first set of features.)
“and deploying a hybrid classifier enabled to classify an input based on a first classification determined, by the first supervised ML classifier, based on the input, and a second classification generated, by the LLM, based on a prompt comprising the input and at least a portion of the LLM taxonomy” (Column 9, Lines 54-59 recite the first classification from the supervised industry classifier, and a second classification from a new industry classifier are generated from the same input data and combined into a final output (hybrid classifier).)
Shacham does not appear to explicitly disclose:
“determining a large language model (LLM) taxonomy comprising definitions for a second subset of the set of nodes;”
“and deploying a hybrid classifier enabled to classify an input based on a first classification determined, by the first supervised ML classifier, based on the input, and a second classification generated, by the LLM, based on a prompt comprising the input and at least a portion of the LLM taxonomy”
However, Wang teaches:
“determining a large language model (LLM) taxonomy comprising definitions for a second subset of the set of nodes;” (Page 9, Section “Discussion” recites leveraging the python dictionary structure to harness the capabilities of the GPT classifiers (zero-shot and few-shot) to execute hierarchical item categorization (determining a large language model (LLM) taxonomy). Page 8, Section “Study 2: Empirical Study” recites level 2 and level 3 activities with distinct common definition, including the utilization of zero-shot classification with only a few succinct descriptive explanations, meaning there is an existence of a separate subset of nodes other than the first subset (level 2, level 3, etc.), such as a second subset of a set of nodes.)
“and deploying a hybrid classifier enabled to classify an input based on a first classification determined, by the first supervised ML classifier, based on the input, and a second classification generated, by the LLM, based on a prompt comprising the input and at least a portion of the LLM taxonomy” (Page 5, Section “Zero-Shot Classifier and Few-Shot Classifier” recites classifiers are equipped to make well-informed predictions, identifying the most likely category for a given input" via "input data into a semantic space, meaning the LLM classifier generates a classification based on inputs. Page 8, Section “Study 2: Empirical Study” further recites a comprehensive input is provided to the GPT-4 LLM. Page 9, Section “Discussion” clarifies the prompt includes an input and a portion of the LLM taxonomy.)
It would have been obvious to a person having ordinary skill in the art (PHOSITA) before the effective filing date of the present application to combine the teachings of Shacham and the teachings of Wang, which are both in the same field of invention. A PHOSITA would be motivated to combine the supervised machine learning classifier architecture from Shacham with the zero-shot classification technique from Wang in order to improve the efficiency of classifying items and solve a deficiency called the “cold start” problem where insufficient training data exists for newer taxonomy categories (see Column 6, Lines 33-39 from Shacham). Thus, implementing the zero-shot classification technique from Wang would improve efficiency by relying on natural language definitions instead of task-specific training data to classify items into hierarchical taxonomy categories This would decrease the amount of resources required to source the training data by eliminating the need to gather third-party references.
With respect to Claim(s) 3:
Shacham and Wang combined teach:
“wherein the hybrid classifier classifies the input by: providing the input to the first supervised ML classifier;” (Column 9, Lines 42-44 from Shacham recite the candidate company may be passed to both the existing industry classifier and the new industry classifier.)
“receiving, from the first supervised ML classifier, the first classification and a first confidence score associated with the first classification;” (Column 9, Lines 44-46 from Shacham recite the existing industry classifier produces one or more existing industry predictions for the candidate company, although a first confidence score is not outputted. Column 9, Lines 60-67 from Shacham recite one or both of the classifiers may also output a confidence score for each of the one or more industry predictions, respectively, that it outputs. This confidence score reflects the corresponding classifier's confidence level in the prediction.)
“providing, to the LLM, the prompt comprising the input and a first portion of the LLM taxonomy;” (Page 8, Section “Study 2: Empirical Study” from Wang recites the amalgamation of the text from the question stem with that from the answer key to provide a comprehensive input for GPT-4 (providing, to the LLM, the prompt comprising the input) and effortlessly employ a descriptive definition and a zero-shot classifier, akin to a first portion of the LLM taxonomy.)
“receiving, from the LLM, the second classification and a second confidence score associated with the second classification;” (Page 8, Section “Study 2: Empirical Study” from Wang recites the amalgamation of the text from the question stem with that from the answer key to provide a comprehensive input for GPT-4 (providing, to the LLM, the prompt comprising the input) and effortlessly employ a descriptive definition and a zero-shot classifier, akin to a first portion of the LLM taxonomy.) Column 9, Lines 44-46 from Shacham recite the existing industry classifier produces one or more existing industry predictions for the candidate company, although a first confidence score is not outputted. Column 9, Lines 60-67 from Shacham recite one or both of the classifiers may also output a confidence score for each of the one or more industry predictions, respectively, that it outputs. This confidence score reflects the corresponding classifier's confidence level in the prediction.)
“and determining a first level classification of the input in the first level of the first taxonomy based on the first output, the second output, the first confidence score, and the second confidence score.” (Column 9, Lines 54-59 from Shacham recite the one or more existing industry predictions and the one or more new industry predictions may be passed to an industry selection component that selects from among the predictions to output one or more final selected industries. Page 7, Section “Study 1: Artificial Data Example” from Wang recites the zero-shot GPT classifier is employed in a sequential manner, utilizing it three times consecutively from higher level (level 1) to lower levels, to categorize words across three nested levels. This is akin to a classification into a level of taxonomy is determined by the output of the LLM classifier operating at the same level. Column 9, Lines 60-67 from Shacham further recite one or both of the classifiers may also output a confidence score for each of the one or more industry predictions, respectively, that it outputs. This confidence score reflects the corresponding classifier's confidence level in the prediction.)
With respect to Claim(s) 5:
Shacham and Wang combined teach:
“wherein determining the LLM taxonomy comprises at least one of: receiving, from a user via a user interface, a definition for a node of the second subset; determining a definition for a node of the second subset based on an analysis of the training data associated with the first subset; or determining a definition for a node of the second subset based on information associated with ancestor nodes of the node of the second subset.” (Column 8, Lines 1-22 from Shacham recites determining information such as a label or characterization of a new-industry (second subnet) node by analyzing information connected to an already existing industry (first subset) node’s training data, akin to determining a definition for a node of the second subset based on an analysis of the training data associated with the first subset.)
With respect to Claim(s) 6, 13, and 20:
Shacham and Wang combined teach:
“wherein the training data comprises at least one of: user-provided training data associated with a node of the first taxonomy; or training data retrieved based on a user-provided document identifier and associated with a node of the first taxonomy, the user-provided document identifier comprising at least one of: a publication identifier, or a patent number.” (Column 7, Lines 40-54 from Shacham recite the training data (profile for Company XYZ) is collected from a users’ own profile submissions, where the user provides a type of industry when creating their profile, akin to user-provided training data associated with a node of the first taxonomy.)
With respect to Claim(s) 8 and 15:
Shacham teaches:
“A system for classifying an input into a node of a taxonomy, the system comprising:
a processor; and
a memory device that stores program code structured to cause the processor to:
receive the input;” (Column 11, Lines 34-38 recite the candidate company is fed into an existing industry classifier, meaning an input was first received before being provided into a classifier.)
“provide the input to a first supervised machine learning (ML) classifier trained using training data to classify data into a first level of the first taxonomy;” (Column 6, Lines 40-42 recite that the existing industry classifier is trained via a supervised machine learning algorithm. (training a first supervised ML classifier based on the training data). Column 6, Lines 50-52 recite the labeled training data fed into the first machine learning algorithm (supervised) consists of a first set of features.)
“receive, from the first supervised ML classifier, a first classification of the input;” (Column 9, Lines 54-59 recite the first classification from the supervised industry classifier, and a second classification from a new industry classifier are generated from the same input data and combined into a final output (hybrid classifier).)
“and determine a first level classification of the input in the first level of the first taxonomy based on the first classification and the second classification.” (Column 9, Lines 54-59 recite the first classification from the supervised industry classifier, and a second classification from a new industry classifier are generated from the same input data and combined into a final output (hybrid classifier).)
Shacham does not appear to explicitly disclose:
“provide, to a large language model (LLM), a first prompt comprising the input and a first portion of an LLM taxonomy comprising definitions for nodes of the first taxonomy;”
“receive, from the LLM, a second classification of the input determined by the LLM based at least on the input and the first portion of the LLM taxonomy;”
However, Wang teaches:
“provide, to a large language model (LLM), a first prompt comprising the input and a first portion of an LLM taxonomy comprising definitions for nodes of the first taxonomy;” (Page 9, Section “Discussion” recites leveraging the python dictionary structure to harness the capabilities of the GPT classifiers (zero-shot and few-shot) to execute hierarchical item categorization (determining a large language model (LLM) taxonomy). Page 8, Section “Study 2: Empirical Study” recites level 2 and level 3 activities with distinct common definition, including the utilization of zero-shot classification with only a few succinct descriptive explanations, meaning there is an existence of a separate subset of nodes other than the first subset (level 2, level 3, etc.), such as a second subset of a set of nodes.)
“receive, from the LLM, a second classification of the input determined by the LLM based at least on the input and the first portion of the LLM taxonomy;” (Page 5, Section “Zero-Shot Classifier and Few-Shot Classifier” recites classifiers are equipped to make well-informed predictions, identifying the most likely category for a given input" via "input data into a semantic space, meaning the LLM classifier generates a classification based on inputs. Page 8, Section “Study 2: Empirical Study” further recites a comprehensive input is provided to the GPT-4 LLM. Page 9, Section “Discussion” clarifies the prompt includes an input and a portion of the LLM taxonomy.)
It would have been obvious to a person having ordinary skill in the art (PHOSITA) before the effective filing date of the present application to combine the teachings of Shacham and the teachings of Wang, which are both in the same field of invention. A PHOSITA would be motivated to combine the supervised machine learning classifier architecture from Shacham with the zero-shot classification technique from Wang in order to solve a deficiency called the “cold start” problem where insufficient training data exists for newer taxonomy categories (see Column 6, Lines 33-39 from Shacham). Thus, implementing the zero-shot classification technique from Wang would address this deficiency by relying on natural language definitions instead of task-specific training data to classify items into hierarchical taxonomy categories.
With respect to Claim(s) 10 and 17:
Shacham and Wang combined teach:
“wherein, to determine a first level classification of the input in the first level of the first taxonomy based on the first classification and the second classification, the program code is structured to further cause the processor to: receive, from the first supervised ML classifier, a first confidence score associated with the first classification;” (Column 9, Lines 44-46 from Shacham recite the existing industry classifier produces one or more existing industry predictions for the candidate company, although a first confidence score is not outputted. Column 9, Lines 60-67 from Shacham recite one or both of the classifiers may also output a confidence score for each of the one or more industry predictions, respectively, that it outputs. This confidence score reflects the corresponding classifier's confidence level in the prediction.)
“receive, from the LLM, a second confidence score associated with the second classification;” (Page 8, Section “Study 2: Empirical Study” from Wang recites the amalgamation of the text from the question stem with that from the answer key to provide a comprehensive input for GPT-4 (providing, to the LLM, the prompt comprising the input) and effortlessly employ a descriptive definition and a zero-shot classifier, akin to a first portion of the LLM taxonomy.) Column 9, Lines 44-46 from Shacham recite the existing industry classifier produces one or more existing industry predictions for the candidate company, although a first confidence score is not outputted. Column 9, Lines 60-67 from Shacham recite one or both of the classifiers may also output a confidence score for each of the one or more industry predictions, respectively, that it outputs. This confidence score reflects the corresponding classifier's confidence level in the prediction.)
“and determine the first level classification of the input in the first level of the first taxonomy based on the first classification, the second classification, the first confidence score, and the second confidence score.” (Column 9, Lines 54-59 from Shacham recite the one or more existing industry predictions and the one or more new industry predictions may be passed to an industry selection component that selects from among the predictions to output one or more final selected industries. Page 7, Section “Study 1: Artificial Data Example” from Wang recites the zero-shot GPT classifier is employed in a sequential manner, utilizing it three times consecutively from higher level (level 1) to lower levels, to categorize words across three nested levels. This is akin to a classification into a level of taxonomy is determined by the output of the LLM classifier operating at the same level. Column 9, Lines 60-67 from Shacham further recite one or both of the classifiers may also output a confidence score for each of the one or more industry predictions, respectively, that it outputs. This confidence score reflects the corresponding classifier's confidence level in the prediction.)
With respect to Claim(s) 12 and 19:
Shacham and Wang combined teach:
“wherein the LLM taxonomy comprises at least one of: a definition for a node of the first taxonomy received from a user via a user interface; a definition for a node of the first taxonomy determined based on an analysis of training data; a definition for a node of the first taxonomy determined based on information associated with subordinate nodes of the node; or a definition for a node of the first taxonomy determined based on information associated with ancestor nodes of the node.” (Page 8, Section “Study 2: Empirical Study” from Wang recites utilizing 10 questions for each domain as prototypical examples (training data) to train GPT-4, meaning a node’s characterization used by the LLM classifier is determined based on an analysis of training data, akin to a definition for a node of the first taxonomy determined based on an analysis of training data.)
Claim(s) 2, 4, 9, 11, 16, and 18 are rejected under 35 U.S.C. 103 as being unpatentable over Shacham et al., (US Patent Application Number US10726355B2 filed on May 31, 2016, hereinafter “Shacham”), in view of “Blueprinting the Future: Automatic Item Categorization Using Hierarchical Zero-Shot and Few-Shot Classifiers” by Wang et al., (non-patent literature published on December 6, 2023, hereinafter “Wang”), in further view of “Stacked Generalization” by David H. Wolpert, (non-patent literature published on February 5, 1992, hereinafter “Wolpert”)
With respect to Claim(s) 2:
Shacham and Wang combined teach:
“wherein the hybrid classifier classifies the input by: providing the input to the first supervised ML classifier;” (Column 9, Lines 42-44 from Shacham recite the candidate company may be passed to both the existing industry classifier and the new industry classifier).
“receiving, from the first supervised ML classifier, the first classification;” (Column 9, Lines 44-46 from Shacham recite the existing industry classifier produces one or more existing industry predictions for the candidate company.)
“receiving, from the LLM, the second classification;” (Page 5, Section “Zero-Shot Classifier and Few-Shot Classifier” from Wang recites the GPT-4 zero-shot classifiers output predictions, which identifies the most likely category for a given input.)
“and determining a first level classification of the input in the first level of the first taxonomy based on the second classification” (Column 9, Lines 54-59 from Shacham recite the one or more existing industry predictions and the one or more new industry predictions may be passed to an industry selection component that selects from among the predictions to output one or more final selected industries. Page 7, Section “Study 1: Artificial Data Example” from Wang recites the zero-shot GPT classifier is employed in a sequential manner, utilizing it three times consecutively from higher level (level 1) to lower levels, to categorize words across three nested levels. This is akin to a classification into a level of taxonomy is determined by the output of the LLM classifier operating at the same level.)
Shacham and Wang combined do not appear to explicitly disclose:
“providing, to the LLM, the prompt comprising the input, a first portion of the LLM taxonomy, and the first classification;”
However, Wolpert teaches:
“providing, to the LLM, the prompt comprising the input, a first portion of the LLM taxonomy, and the first classification;” (Page 11, Section “iv” recites that the guesses of a first-level classifier (first classification) are used as components of the input provided to a second-level generalizer, alongside information about the original question, akin to a prompt (the input provided to the second-stage generalizer) that comprises the input and the first classification.)
It would have been obvious to a person having ordinary skill in the art (PHOSITA) before the effective filing date of the present application to combine the teachings of Shacham and the teachings of Wang with the teachings of Wolpert, which are all in the same field of invention. A PHOSITA would be motivated to combine the supervised machine learning classifier architecture and zero-shot classification technique from Shacham and Wang with the stacked generalization technique from Wolpert, which would improve classification accuracy by allowing the LLM classifier to account for the supervised ML classifier’s prediction when making its own determination.
With respect to Claim(s) 4:
Shacham and Wang combined teach:
“training a second supervised ML classifier based on the training data, the second supervised ML classifier trained to classify data into a second level of the supervised ML taxonomy;” (Column 6, Lines 4-9 from Shacham recite two separate supervised ML classifiers, where the second is known as the new industry classifier. Each of these classifiers are trained separately based on training data. Column 6, Lines 40-50 from Shacham further clarifies the new industry classifier is trained by a second machine learning algorithm to classify a candidate company into a new industry, where labels may be provided for training data indicating appropriate industries for companies in the training data, and the second set of features for the labeled training data are fed to the second machine learning algorithm to train the new industry classifier. Page 7, Section “Study 1: Artificial Data Example” from Wang recites the zero-shot GPT classifier is employed in a sequential manner, utilizing it three times consecutively from higher level (level 1) to lower levels, to categorize words across three nested levels.)
“receiving, from the second supervised ML classifier, a third classification of the input;” (Column 9. Lines 45-51 from Shacham recite the new industry classifier (second supervised ML classifier) produces one or more new industry predictions for the candidate, meaning a classification output is received from a second classifier for a given input, akin to receiving a third classification of the input.)
“providing, to a large language model (LLM), a second prompt comprising the input, the first level classification, and a second portion of the LLM taxonomy;” (Page 7, Section “Study 1: Artificial Data Example” from Wang recites the zero-shot GPT classifier (LLM) is employed in a sequential manner, utilizing it three times consecutively from higher level (level 1) to lower levels (separate prompts, including a second prompt), to categorize words across three nested levels. This is akin to a classification into a second level of taxonomy is determined by the output of the LLM classifier operating at the same level.)
“receiving, from the LLM, a fourth classification determined by the LLM based at least on the input, the first level classification, and the second portion of the LLM taxonomy;” (Column 9. Lines 45-51 from Shacham recite the new industry classifier (second supervised ML classifier) produces one or more new industry predictions for the candidate, meaning multiple classification outputs can be received from a second classifier for a given input, akin to receiving a fourth classification of the input. Page 7, Section “Study 1: Artificial Data Example” from Wang recites the zero-shot GPT classifier (LLM) is employed in a sequential manner, utilizing it three times consecutively from higher level (level 1) to lower levels (separate prompts, including a second prompt), to categorize words across three nested levels. This is akin to a classification into a second level of taxonomy is determined by the output of the LLM classifier operating at the same level.)
“determining a second level classification of the input in the second level of the first taxonomy based on the third classification and the fourth classification, the second level of the first taxonomy being subordinate to the first level of the first taxonomy;” (Page 8, Section “Study 2: Empirical Study” from Wang recites sequential classification produces a distinct classification output specifically corresponding to the second taxonomy level (called level 2). Column 9, Lines 54-59 from Shacham recite the one or more existing industry predictions and the one or more new industry predictions may be passed to an industry selection component that selects from among the predictions to output one or more final selected industries, meaning a final classification is determined by considering two separate classifier outputs together.)
“and providing, as an overall classification of the input, an output classification comprising the first level classification and the second level classification.” (Page 17, Appendix Figure from Wang shows the output classification result modeled through a program, where it shows the output classification is comprised of classifications from many distinct taxonomy levels combined together into a single output, including the processed Category 1 (first level classification) and the processed Category 2 (second level classification).)
Shacham and Wang combined do not appear to explicitly disclose:
“providing, to the second supervised ML classifier, the input and the first level classification;”
“providing, to a large language model (LLM), a second prompt comprising the input, the first level classification, and a second portion of the LLM taxonomy;”
However, Wolpert teaches:
“providing, to the second supervised ML classifier, the input and the first level classification;” (Page 11, Section “iv” recites that the guesses of a first-level classifier (first classification) are used as components of the input provided to a second-level generalizer, alongside information about the original question, akin to a prompt (the input provided to the second-stage generalizer) that comprises the input and the first classification.)
“providing, to a large language model (LLM), a second prompt comprising the input, the first level classification, and a second portion of the LLM taxonomy;” (Page 11, Section “iv” recites that the guesses of a first-level classifier (first classification) are used as components of the input provided to a second-level generalizer, alongside information about the original question, akin to a prompt (the input provided to the second-stage generalizer) that comprises the first classification.)
It would have been obvious to a person having ordinary skill in the art (PHOSITA) before the effective filing date of the claimed invention to combine the supervised ML classifier training techniques and LLM classification techniques of Shacham and Wang with the stacked generalization technique of Wolpert since all three are directed to methods of training and combining classifiers to improve classification accuracy. A PHOSITA would have been motivated to apply Wolpert's teaching of providing a prior classifier's output as an input to a subsequent classifier or generalizer to the overall system, including at the second taxonomy level, in order to reduce generalization errors and improve classification accuracy through a stacking approach.
With respect to Claim(s) 9 and 16:
Shacham and Wang combined teach:
“and determine the second classification as the first level classification of the input in the first level of the first taxonomy.” (Column 9, Lines 54-59 from Shacham recite the one or more existing industry predictions and the one or more new industry predictions may be passed to an industry selection component that selects from among the predictions to output one or more final selected industries. Page 7, Section “Study 1: Artificial Data Example” from Wang recites the zero-shot GPT classifier is employed in a sequential manner, utilizing it three times consecutively from higher level (level 1) to lower levels, to categorize words across three nested levels. This is akin to a classification into a level of taxonomy is determined by the output of the LLM classifier operating at the same level.)
Shacham and Wang combined do not appear to explicitly disclose:
“wherein, to determine a first level classification of the input in the first level of the first taxonomy based on the first classification and the second classification, the program code is structured to further cause the processor to: provide the first classification in the prompt to the LLM, the LLM determining the second classification further based on the first classification;”
However, Wolpert teaches:
“wherein, to determine a first level classification of the input in the first level of the first taxonomy based on the first classification and the second classification, the program code is structured to further cause the processor to: provide the first classification in the prompt to the LLM, the LLM determining the second classification further based on the first classification;” (Page 11, Section “iv” recites that the guesses of a first-level classifier (first classification) are used as components of the input provided to a second-level generalizer, alongside information about the original question, akin to a prompt (the input provided to the second-stage generalizer) that comprises the input and the first classification.)
It would have been obvious to a person having ordinary skill in the art (PHOSITA) before the effective filing date of the present application to combine the teachings of Shacham and the teachings of Wang with the teachings of Wolpert, which are all in the same field of invention. A PHOSITA would be motivated to combine the supervised machine learning classifier architecture and zero-shot classification technique from Shacham and Wang with the stacked generalization technique from Wolpert, which would improve classification accuracy by allowing the LLM classifier to account for the supervised ML classifier’s prediction when making its own determination.
With respect to Claim(s) 11 and 18:
Shacham and Wang combined teach:
“wherein the program code is structured to further cause the processor to: provide the input and the first level classification to a second supervised ML classifier trained using the training data to classify data into a second level of the first taxonomy that is subordinate to the first level;” (Column 6, Lines 4-9 from Shacham recite two separate supervised ML classifiers, where the second is known as the new industry classifier. Each of these classifiers are trained separately based on training data. Column 6, Lines 40-50 from Shacham further clarifies the new industry classifier is trained by a second machine learning algorithm to classify a candidate company into a new industry, where labels may be provided for training data indicating appropriate industries for companies in the training data, and the second set of features for the labeled training data are fed to the second machine learning algorithm to train the new industry classifier. Page 7, Section “Study 1: Artificial Data Example” from Wang recites the zero-shot GPT classifier is employed in a sequential manner, utilizing it three times consecutively from higher level (level 1) to lower levels, to categorize words across three nested levels.)
“receive, from the second supervised ML classifier, a third classification for the input;” (Column 9. Lines 45-51 from Shacham recite the new industry classifier (second supervised ML classifier) produces one or more new industry predictions for the candidate, meaning a classification output is received from a second classifier for a given input, akin to receiving a third classification of the input.)
“provide, to a large language model (LLM), a second prompt comprising the input, the first level classification, and a second portion of the LLM taxonomy;” (Page 7, Section “Study 1: Artificial Data Example” from Wang recites the zero-shot GPT classifier (LLM) is employed in a sequential manner, utilizing it three times consecutively from higher level (level 1) to lower levels (separate prompts, including a second prompt), to categorize words across three nested levels. This is akin to a classification into a second level of taxonomy is determined by the output of the LLM classifier operating at the same level.)
“receive, from the LLM, a fourth classification, the fourth classification determined by the LLM based at least on the input, the first level classification, and the second portion of the LLM taxonomy;” (Column 9. Lines 45-51 from Shacham recite the new industry classifier (second supervised ML classifier) produces one or more new industry predictions for the candidate, meaning multiple classification outputs can be received from a second classifier for a given input, akin to receiving a fourth classification of the input. Page 7, Section “Study 1: Artificial Data Example” from Wang recites the zero-shot GPT classifier (LLM) is employed in a sequential manner, utilizing it three times consecutively from higher level (level 1) to lower levels (separate prompts, including a second prompt), to categorize words across three nested levels. This is akin to a classification into a second level of taxonomy is determined by the output of the LLM classifier operating at the same level.)
“determine a second level classification of the input in a second level of the first taxonomy based on the third classification and the fourth classification, the second level of the first taxonomy being subordinate to the first level of the first taxonomy;” (Page 8, Section “Study 2: Empirical Study” from Wang recites sequential classification produces a distinct classification output specifically corresponding to the second taxonomy level (called level 2). Column 9, Lines 54-59 from Shacham recite the one or more existing industry predictions and the one or more new industry predictions may be passed to an industry selection component that selects from among the predictions to output one or more final selected industries, meaning a final classification is determined by considering two separate classifier outputs together.)
“and provide, as an overall classification of the input, an output classification comprising the first level classification and the second level classification.” (Page 17, Appendix Figure from Wang shows the output classification result modeled through a program, where it shows the output classification is comprised of classifications from many distinct taxonomy levels combined together into a single output, including the processed Category 1 (first level classification) and the processed Category 2 (second level classification).)
Shacham and Wang combined do not appear to explicitly disclose:
“provide, to a large language model (LLM), a second prompt comprising the input, the first level classification, and a second portion of the LLM taxonomy;”
However, Wolpert teaches:
“provide, to a large language model (LLM), a second prompt comprising the input, the first level classification, and a second portion of the LLM taxonomy;” (Page 11, Section “iv” recites that the guesses of a first-level classifier (first classification) are used as components of the input provided to a second-level generalizer, alongside information about the original question, akin to a prompt (the input provided to the second-stage generalizer) that comprises the first classification.)
It would have been obvious to a person having ordinary skill in the art (PHOSITA) before the effective filing date of the claimed invention to combine the supervised ML classifier training techniques and LLM classification techniques of Shacham and Wang with the stacked generalization technique of Wolpert since all three are directed to methods of training and combining classifiers to improve classification accuracy. A PHOSITA would have been motivated to apply Wolpert's teaching of providing a prior classifier's output as an input to a subsequent classifier or generalizer to the overall system, including at the second taxonomy level, in order to reduce generalization errors and improve classification accuracy through a stacking approach.
Claim(s) 7 and 14 are rejected under 35 U.S.C. 103 as being unpatentable over Shacham et al., (US Patent Application Number US10726355B2 filed on May 31, 2016, hereinafter “Shacham”), in view of “Blueprinting the Future: Automatic Item Categorization Using Hierarchical Zero-Shot and Few-Shot Classifiers” by Wang et al., (non-patent literature published on December 6, 2023, hereinafter “Wang”), in further view of Schmidtler et al., (US Patent Application Number US8239335B2 filed on February 23, 2011, hereinafter “Schmidtler”)
With respect to Claim(s) 7:
Shacham and Wang combined do not appear to explicitly disclose:
“wherein the first taxonomy is a hierarchical taxonomy, and determining the first taxonomy comprises at least one of: receiving, from a user via a user interface, at least a portion of the first taxonomy; automatically generating at least a portion of the first taxonomy based on user input; or generating the first taxonomy based on user modifications to a previously generated taxonomy.”
However, Schmidtler teaches:
“wherein the first taxonomy is a hierarchical taxonomy, and determining the first taxonomy comprises at least one of: receiving, from a user via a user interface, at least a portion of the first taxonomy; automatically generating at least a portion of the first taxonomy based on user input; or generating the first taxonomy based on user modifications to a previously generated taxonomy.” (Column 6, Lines 57-61 describe an existing branch of the taxonomy consisting of nodes. It is understood that this taxonomy is hierarchical due to the branch and node structure. Column 28, Lines 64-67 and Column 29, Lines 1-5 recite that when an unlabeled document’s classification confidence falls below a threshold during manual-review from a user input, a document with a certain confidence level triggers the creation of a new category that is added to the system, where the document is then assigned to the new category. This is akin to automatically generating at least a portion of the first taxonomy based on user input.)
It would have been obvious to a person having ordinary skill in the art (PHOSITA) before the effective filing date of the present application to combine the teachings of Shacham and the teachings of Wang with the teachings of Schmidtler, which are all in the same field of invention. A PHOSITA would be motivated to combine the supervised machine learning classifier architecture and zero-shot classification technique from Shacham and Wang with Schmidtler, which routes low-confidence classifications to a user interface for review and uses the user’s input to modify the taxonomy, in order to improve classification and avoid misclassification through automated mechanisms that signal human review is required.
With respect to Claim(s) 14:
Shacham and Wang combined do not appear to explicitly disclose:
“wherein the first taxonomy comprises at least one of: a hierarchical taxonomy received from a user via a user interface; a hierarchical taxonomy automatically generate based on user input; or a hierarchical taxonomy generated based on user modifications to a previously generated taxonomy.”
However, Schmidtler teaches:
“wherein the first taxonomy comprises at least one of: a hierarchical taxonomy received from a user via a user interface; a hierarchical taxonomy automatically generate based on user input; or a hierarchical taxonomy generated based on user modifications to a previously generated taxonomy.” (Column 6, Lines 57-61 describe an existing branch of the taxonomy consisting of nodes. It is understood that this taxonomy is hierarchical due to the branch and node structure. Column 28, Lines 64-67 and Column 29, Lines 1-5 recite that when an unlabeled document’s classification confidence falls below a threshold during manual-review from a user input, a document with a certain confidence level triggers the creation of a new category that is added to the system, where the document is then assigned to the new category. This is akin to automatically generating at least a portion of the first taxonomy based on user input.)
It would have been obvious to a person having ordinary skill in the art (PHOSITA) before the effective filing date of the present application to combine the teachings of Shacham and the teachings of Wang with the teachings of Schmidtler, which are all in the same field of invention. A PHOSITA would be motivated to combine the supervised machine learning classifier architecture and zero-shot classification technique from Shacham and Wang with Schmidtler, which routes low-confidence classifications to a user interface for review and uses the user’s input to modify the taxonomy, in order to improve classification and avoid misclassification through automated mechanisms that signal human review is required.
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to Vibha Bhat whose telephone number is (571)-272-7091. The examiner can normally be reached on Monday – Thursday from 8:00 AM to 5:00 PM EST and every other Friday from 8:00 AM to 4:00 PM EST.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. See MPEP § 713.01. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at https://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Mariela Reyes, can be reached at telephone number (571)-270-1006. The fax phone number for the organization where this application or proceeding is assigned is (571)-273-8300. Information regarding the status of an application may be obtained from 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://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 (572)-272-1000.
/Vibha Bhat/Examiner
Art Unit 2142
/HAIMEI JIANG/Primary Examiner, Art Unit 2142
Prosecution Insights