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 .
This action is in response to the amendment filed on Jan. 2nd, 2026. The amendments are linked to the original application filed on Dec. 19th, 2022.
Response to Amendment
The Examiner thanks the applicant for the remarks, edits and arguments.
Regarding Drawing Objections
Applicant Remarks:
The applicant has amended the specification to correct a naming error. Further the applicant has not added any new subject matter and has made amendments to comply with 37 CFR §1.84(p)(4) and request withdrawal of the objection.
Examiner Response:
The examiner has reviewed the specification and the drawing and find that they currently comply with 37 CFR §1.84(p)(4). Therefore, the examiner is withdrawing the drawing objection under 37 CFR §1.84(p)(4).
Regarding Claim Rejections – 35 U.S.C. §101
Applicant Remarks:
The applicant has made amendments to the independent claims to provide further explanation of the claims and to no longer recite abstract ideas. The claimed subject matter discloses the use of machine learning models and use of online databases. The applicant has made amendments to further disclose the claimed invention is integrated into a practical application and provides an improvement to machine learning systems. The applicant points to the specification which states the proposed improvements of the system. In light of the specification, the applicant believes the current amened claims can be integrated into a practical application and would comply with 35 U.S.C. 101. Therefore, the applicant requests the rejection under 35 U.S.C. 101 be withdrawn.
Examiner Response:
The examiner has noted the amendments to the claims and, as required, the examiner must apply the Alice/Mayo test to the amended claims to determine patent subject matter eligibility. The applicant argues that the amended claims recite additional limitations which integrate the alleged abstract ideas into practical applications, “specifically an improvement to providing explanations for live transactions in a database Management System.”. The examiner believes the claims and specification fail to demonstrate to one of ordinary skill in the art a technical improvement to a computing device or technical field. The process recited in the specification still requires the user to generate initial explanations and, as seen in Fig. 5 540, if a precomputed explanation is not located, the system will generate a new explanation and therefore, cause the system to perform more computations to produce the same result as other XAI models.
Next, the examiner must evaluate the additional limitations to see if they integrate into a practical application as stated in MPEP 2106.04 (d)(II), “ Examiners evaluate integration into a practical application by: (1) identifying whether there are any additional elements recited in the claim beyond the judicial exception(s); and (2) evaluating those additional elements individually and in combination to determine whether they integrate the exception into a practical application, using one or more of the considerations introduced in subsection I supra, and discussed in more detail in MPEP §§ 2106.04(d)(1), 2106.04(d)(2), 2106.05(a) through (c) and 2106.05(e) through (h).”. The examiner believes, based on the MPEP, the additional amended limitations fail to integrate into a practical application. In particular the independent claims recite, “generating a unified explainability score for each input space cluster in the homogeneous region based on the explainability scores; and”. The examiner believes this limitation falls under 2106.05(f) Mere Instructions To Apply An Exception, specifically 2106.05(f)(2). The examiner believes the limitation recites the use of a computer as a tool to perform an existing process of applying a value to a cluster based on an evaluation. Next the examiner believes, “storing each unified explainability score in a set of precomputed explanations; and”, from the independent claims falls under 2106.05(g) and recites an insignificant extra-activity solution, specifically it recites Well-Understood, Routine, Conventual activities. The examiner believes this recites a process of storing and retrieving information from memory. Further evaluation of the additional elements can be found below under 101 Rejection.
Finally, the examiner has taken into consideration the amended claims and the specification and evaluated the claims using the Alice/Mayo test for patent eligibility. The examiner believes, for the reasons stated above, the current amended claims recite ineligible subject matter. Therefore, the rejection under U.S.C. 101 is upheld, see 101 rejection below.
Regarding Claim Rejections – 35 U.S.C. §103
Applicant Remarks:
The applicant has amended the claims to further distinguish the claimed subject matter from prior arts proposed. The applicant states the combination of Morichetta and Galitsky fails to teach the amended claims. In particular the applicant states these arts fail to disclose hierarchical clustering of data from databases to produce the claimed clustered structure. Next the applicant states the arts fail to disclose the scoring process or matching labels after scoring as disclosed in the claims and the specification.
Finally, the applicant believes the combination of proposed arts fail to teach the independent claims. Therefore, the art combinations used to teach the dependent claims would be considered moot by virtue of dependency. For the reasons in the submitted remarks and reasons stated above, the applicant believes that the current amended claims art not taught by the combination of proposed arts and therefore request the rejection under 35 U.S.C. 103 be withdrawn.
Examiner Response:
The applicant states that the combination of Morichetta and Galitsky fail to disclose the amended subject matter. After each amendment, the examiner must evaluate the amended claims and review the prior art to see if it is still applicable. Further, the examiner must perform a complete search of the amended claims and limitations. While evaluating the previously proposed arts, the examiner noted the arts, Morichetta and Galitsky disclose similar elements of the claimed invention but fail to explicitly disclose the amended limitations. However, while performing a complete search, the examiner has found new subject matter that is able to disclose the amended claims. Therefore, the examiner no longer relies on Morichetta and Galitsky to teach the claimed subject matter and instead relies on new art. The examiner believes it would be obvious for a person of ordinary skill in the art would be able to combine the arts Zhao, previously presented, and Dalli et al (US 20220398460 A1), to disclose the claimed subject matter. Finally, the Zhao and Dalli are used in combination to disclose the independent claims and the dependent claims. Therefore, the rejection under U.S.C. 103 is upheld, see 103 rejection below.
Claim Rejections - 35 USC § 101
35 U.S.C. 101 reads as follows:
Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.
Claims 1, 2, 4-12, and 14-22 are rejected under 35 U.S.C 101 because the claimed invention is directed to an abstract idea without significantly more. The analysis of the claims will follow the 2019 Revised Patent Subject Matter Eligibility Guidance, 84 Fed. Reg. 50 (“2019 PEG”).
Claim 1
Step 1 – Is the claim to a process, machine, manufacture or composition of matter?
Claim 1 recites, “A method of outputting explanations for live transactions implemented by a database management system, the database management system comprising a processor communicatively coupled to a memory, and the method comprising:” therefore it is directed to the statutory category of a process.
Step 2A Prong 1 – Does the claim recite an abstract idea, law of nature, or natural phenomenon?
The claim recites, inter alia:
“hierarchically clustering an input space of the labeled transactions based on the input features to form input space clusters of the labeled transactions divided into branch levels;” Under its broadest reasonable interpretation in light of the specification, this limitation encompasses mathematical concept of utilizing a mathematical formula to perform calculations. Hierarchical clustering is a mathematical process that can be completed by a human with pen and paper or using a computer as a tool. This claim discloses a math operation and therefore is ineligible.
“identifying a homogenous region of the input space, wherein the identifying comprises:” Under its broadest reasonable interpretation in light of the specification, this limitation encompasses the mental process of evaluating and observing data, which is an evaluation or observation that is practically capable of being performed in the human mind with the assistance of pen and paper. A human is able to evaluate data and identify a cluster. The limitation is merely applying an abstract idea on generic computer system. See MPEP 2106.04(a)(2)(III)(c).
“scoring homogeneity of the input space clusters based on a number of matching output labels from the output labels corresponding to the labeled transactions in each input space cluster;” Under its broadest reasonable interpretation in light of the specification, this limitation encompasses the mental process of evaluating and observing data, which is an evaluation or observation that is practically capable of being performed in the human mind with the assistance of pen and paper. A human is able to evaluate clusters and provide a value as an opinion or judgment. The limitation is merely applying an abstract idea on generic computer system. See MPEP 2106.04(a)(2)(III)(c).
“scoring homogeneity of the branch levels based on the homogeneity scoring of the input space clusters; and” Under its broadest reasonable interpretation in light of the specification, this limitation encompasses the mental process of evaluating and observing data, which is an evaluation or observation that is practically capable of being performed in the human mind with the assistance of pen and paper. A human is able to evaluate data and apply values to the data based on judgement. The limitation is merely applying an abstract idea on generic computer system. See MPEP 2106.04(a)(2)(III)(c).
“selecting the homogeneous region based on the homogeneity scoring of the branch levels;” Under its broadest reasonable interpretation in light of the specification, this limitation encompasses the mental process of evaluating and observing data, which is an evaluation or observation that is practically capable of being performed in the human mind with the assistance of pen and paper. A human is able to evaluate data and apply a value based on judgement. The limitation is merely applying an abstract idea on generic computer system. See MPEP 2106.04(a)(2)(III)(c).
“selecting, by the processor using the XAI module, and from the set of precomputed explanations, an explainability score for a live transaction.” Under its broadest reasonable interpretation in light of the specification, this limitation encompasses the mental process of evaluating and observing data, which is an evaluation or observation that is practically capable of being performed in the human mind with the assistance of pen and paper. A human is able to evaluate data and make a decision or opinion based on the evaluation. The limitation is merely applying an abstract idea on generic computer system. See MPEP 2106.04(a)(2)(III)(c).
Step 2A Prong 2 – Does the claim recite additional elements that integrate the judicial exception into a practical application?
The claim recites the additional elements, “generating, by the processor, an explainable artificial intelligence (XAI) module, wherein the generating comprises:” amounts to generic computer components used as a tool to perform an existing process. Thus, the additional element amounts to no more than a recitation of the words "apply it" (or an equivalent) or are more than mere instructions to implement an abstract idea or other exception on a computer (see MPEP § 2106.05(f)).
“obtaining a set of labeled transactions completed by the database management system, the labeled transactions comprising input features and corresponding output labels generated by a machine learning (ML) model;” is an insignificant extra-solution activity required for any uses of the mental processes (see MPEP § 2106.05(g)) As such, the claim is ineligible.
“obtaining, in response to the identifying and by an explainability model, explainability scores for labeled transactions in each input space cluster in the homogeneous region;” amounts to generic computer components used as a tool to perform an existing process. Thus, the additional element amounts to no more than a recitation of the words "apply it" (or an equivalent) or are more than mere instructions to implement an abstract idea or other exception on a computer (see MPEP § 2106.05(f)).
“generating a unified explainability score for each input space cluster in the homogeneous region based on the explainability scores; and” amounts to generic computer components used as a tool to perform an existing process. Thus, the additional element amounts to no more than a recitation of the words "apply it" (or an equivalent) or are more than mere instructions to implement an abstract idea or other exception on a computer (see MPEP § 2106.05(f)).
“storing each unified explainability score in a set of precomputed explanations; and” is an insignificant extra-solution activity required for any uses of the mental processes (see MPEP § 2106.05(g)) As such, the claim is ineligible.
Step 2B – Does the claim recite additional elements that amount to significantly more than the judicial exception?
Finally, the claim taken as a whole does not contain an inventive concept which provides significantly more than the abstract idea. The additional elements, “generating, by the processor, an explainable artificial intelligence (XAI) module, wherein the generating comprises:” amounts to generic computer components used as a tool to perform an existing process. Thus, the additional element amounts to no more than a recitation of the words "apply it" (or an equivalent) or are more than mere instructions to implement an abstract idea or other exception on a computer (see MPEP § 2106.05(f)).
“obtaining a set of labeled transactions completed by the database management system, the labeled transactions comprising input features and corresponding output labels generated by a machine learning (ML) model;” is an insignificant extra-solution activity required for any uses of abstract ideas (see MPEP § 2106.05(g)), and is a well-understood, routine, conventional activity (see MPEP § 2106.05(d)(II)(i); “Receiving or transmitting data over a network, e.g., using the Internet to gather data”.
“obtaining, in response to the identifying and by an explainability model, explainability scores for labeled transactions in each input space cluster in the homogeneous region;” amounts to generic computer components used as a tool to perform an existing process. Thus, the additional element amounts to no more than a recitation of the words "apply it" (or an equivalent) or are more than mere instructions to implement an abstract idea or other exception on a computer (see MPEP § 2106.05(f)).
“generating a unified explainability score for each input space cluster in the homogeneous region based on the explainability scores; and” amounts to generic computer components used as a tool to perform an existing process. Thus, the additional element amounts to no more than a recitation of the words "apply it" (or an equivalent) or are more than mere instructions to implement an abstract idea or other exception on a computer (see MPEP § 2106.05(f)).
“storing each unified explainability score in a set of precomputed explanations; and” is an insignificant extra-solution activity required for any uses of abstract ideas (see MPEP § 2106.05(g)), and is a well-understood, routine, conventional activity (see MPEP § 2106.05(d)(iv); “Storing and retrieving information in memory”.
Taken alone or in combination, the additional elements of the claim do not provide an inventive concept and thus the claim is subject-matter ineligible.
Claim 2
Step 1 – Is the claim to a process, machine, manufacture or composition of matter?
A process, as above.
Step 2A Prong 1 – Does the claim recite an abstract idea, law of nature, or natural phenomenon?
The claim recites the abstract ideas of the preceding claims from which it depends.
Step 2A Prong 2 – Does the claim recite additional elements that integrate the judicial exception into a practical application?
The claim recites the additional elements, “receiving a second live transaction from a user device.” is an insignificant extra-solution activity required for any uses of the mental processes (see MPEP § 2106.05(g)) As such, the claim is ineligible.
Step 2B – Does the claim recite additional elements that amount to significantly more than the judicial exception?
Finally, the claim taken as a whole does not contain an inventive concept which provides significantly more than the abstract idea. The additional elements, “receiving a second live transaction from a user device.” is an insignificant extra-solution activity required for any uses of abstract ideas (see MPEP § 2106.05(g)), and is a well-understood, routine, conventional activity (see MPEP § 2106.05(d)(II)(i); “Receiving or transmitting data over a network, e.g., using the Internet to gather data”.
Taken alone or in combination, the additional elements of the claim do not provide an inventive concept and thus the claim is subject-matter ineligible.
Claim 3 (Cancelled)
Claim 4
Step 1 – Is the claim to a process, machine, manufacture or composition of matter?
A process, as above.
Step 2A Prong 1 – Does the claim recite an abstract idea, law of nature, or natural phenomenon?
The claim recites the abstract ideas of the preceding claims from which it depends.
Step 2A Prong 2 – Does the claim recite additional elements that integrate the judicial exception into a practical application?
The claim recites the additional elements, “wherein the explainability score is selected based on input features of the live transaction.” amounts to generic computer components used as a tool to perform an existing process. Thus, the additional element amounts to no more than a recitation of the words "apply it" (or an equivalent) or are more than mere instructions to implement an abstract idea or other exception on a computer (see MPEP § 2106.05(f)).
Step 2B – Does the claim recite additional elements that amount to significantly more than the judicial exception?
Finally, the claim taken as a whole does not contain an inventive concept which provides significantly more than the abstract idea. The additional elements, “wherein the explainability score is selected based on input features of the live transaction.” amounts to generic computer components used as a tool to perform an existing process. Thus, the additional element amounts to no more than a recitation of the words "apply it" (or an equivalent) or are more than mere instructions to implement an abstract idea or other exception on a computer (see MPEP § 2106.05(f)).
Taken alone or in combination, the additional elements of the claim do not provide an inventive concept and thus the claim is subject-matter ineligible.
Claim 5
Step 1 – Is the claim to a process, machine, manufacture or composition of matter?
A process, as above.
Step 2A Prong 1 – Does the claim recite an abstract idea, law of nature, or natural phenomenon?
The claim recites, inter alia:
“selecting, by the XAI module, a most aligned cluster from the homogeneous region based on input features of the second live transaction; and” Under its broadest reasonable interpretation in light of the specification, this limitation encompasses the mental process of evaluating and observing data, which is an evaluation or observation that is practically capable of being performed in the human mind with the assistance of pen and paper. A human is able to evaluate a set of clusters and judge which cluster is the most aligned. The limitation is merely applying an abstract idea on generic computer system. See MPEP 2106.04(a)(2)(III)(c).
“determining, by the XAI module, whether the most aligned cluster satisfies a closeness criterion.” Under its broadest reasonable interpretation in light of the specification, this limitation encompasses the mental process of evaluating and observing data, which is an evaluation or observation that is practically capable of being performed in the human mind with the assistance of pen and paper. A human is able to evaluate a cluster to determine it meet criteria and provide an opinion. The limitation is merely applying an abstract idea on generic computer system. See MPEP 2106.04(a)(2)(III)(c).
Step 2A Prong 2 – Does the claim recite additional elements that integrate the judicial exception into a practical application?
This claim does not recite any additional limitations which integrate the abstract idea into a practical application.
Step 2B – Does the claim recite additional elements that amount to significantly more than the judicial exception?
Finally, the claim taken as a whole does not contain an inventive concept which provides significantly more than the abstract idea and thus the claim is subject-matter ineligible.
Claim 6
Step 1 – Is the claim to a process, machine, manufacture or composition of matter?
A process, as above.
Step 2A Prong 1 – Does the claim recite an abstract idea, law of nature, or natural phenomenon?
The claim recites the abstract ideas of the preceding claims from which it depends.
Step 2A Prong 2 – Does the claim recite additional elements that integrate the judicial exception into a practical application?
The claim recites the additional elements, “in response to determining that the most aligned cluster does not satisfy the closeness criterion, generating a new explainability score for the second live transaction.” amounts to generic computer components used as a tool to perform an existing process. Thus, the additional element amounts to no more than a recitation of the words "apply it" (or an equivalent) or are more than mere instructions to implement an abstract idea or other exception on a computer (see MPEP § 2106.05(f)).
Step 2B – Does the claim recite additional elements that amount to significantly more than the judicial exception?
Finally, the claim taken as a whole does not contain an inventive concept which provides significantly more than the abstract idea. The additional elements, “in response to determining that the most aligned cluster does not satisfy the closeness criterion, generating a new explainability score for the second live transaction.” amounts to generic computer components used as a tool to perform an existing process. Thus, the additional element amounts to no more than a recitation of the words "apply it" (or an equivalent) or are more than mere instructions to implement an abstract idea or other exception on a computer (see MPEP § 2106.05(f)).
Taken alone or in combination, the additional elements of the claim do not provide an inventive concept and thus the claim is subject-matter ineligible.
Claim 7
Step 1 – Is the claim to a process, machine, manufacture or composition of matter?
A process, as above.
Step 2A Prong 1 – Does the claim recite an abstract idea, law of nature, or natural phenomenon?
The claim recites, inter alia:
“in response to determining that the most aligned cluster satisfies the closeness criterion, comparing output labels of the most aligned cluster with output labels of the second live transaction.” Under its broadest reasonable interpretation in light of the specification, this limitation encompasses the mental process of evaluating and observing data, which is an evaluation or observation that is practically capable of being performed in the human mind with the assistance of pen and paper. A human is able to evaluate data and compare labels and provide and opinion or judgement based on the comparison. The limitation is merely applying an abstract idea on generic computer system. See MPEP 2106.04(a)(2)(III)(c).
Step 2A Prong 2 – Does the claim recite additional elements that integrate the judicial exception into a practical application?
This claim does not recite any additional limitations which integrate the abstract idea into a practical application.
Step 2B – Does the claim recite additional elements that amount to significantly more than the judicial exception?
Finally, the claim taken as a whole does not contain an inventive concept which provides significantly more than the abstract idea and thus the claim is subject-matter ineligible.
Claim 8
Step 1 – Is the claim to a process, machine, manufacture or composition of matter?
A process, as above.
Step 2A Prong 1 – Does the claim recite an abstract idea, law of nature, or natural phenomenon?
The claim recites, inter alia:
“determining, based on the comparing, that the output labels do not include a congruent output label; and” Under its broadest reasonable interpretation in light of the specification, this limitation encompasses the mental process of evaluating and observing data, which is an evaluation or observation that is practically capable of being performed in the human mind with the assistance of pen and paper. A human is able to evaluate data and determine if matching labels are present. The limitation is merely applying an abstract idea on generic computer system. See MPEP 2106.04(a)(2)(III)(c).
Step 2A Prong 2 – Does the claim recite additional elements that integrate the judicial exception into a practical application?
The claim recites the additional elements, “in response to the determining that the output labels do not include a congruent output label, generating a new explainability score for the second live transaction.” amounts to generic computer components used as a tool to perform an existing process. Thus, the additional element amounts to no more than a recitation of the words "apply it" (or an equivalent) or are more than mere instructions to implement an abstract idea or other exception on a computer (see MPEP § 2106.05(f)).
Step 2B – Does the claim recite additional elements that amount to significantly more than the judicial exception?
Finally, the claim taken as a whole does not contain an inventive concept which provides significantly more than the abstract idea. The additional elements, “in response to the determining that the output labels do not include a congruent output label, generating a new explainability score for the second live transaction.” amounts to generic computer components used as a tool to perform an existing process. Thus, the additional element amounts to no more than a recitation of the words "apply it" (or an equivalent) or are more than mere instructions to implement an abstract idea or other exception on a computer (see MPEP § 2106.05(f)).
Taken alone or in combination, the additional elements of the claim do not provide an inventive concept and thus the claim is subject-matter ineligible.
Claim 9
Step 1 – Is the claim to a process, machine, manufacture or composition of matter?
A process, as above.
Step 2A Prong 1 – Does the claim recite an abstract idea, law of nature, or natural phenomenon?
The claim recites, inter alia:
“identifying a congruent output label based on the comparing; and” Under its broadest reasonable interpretation in light of the specification, this limitation encompasses the mental process of evaluating and observing data, which is an evaluation or observation that is practically capable of being performed in the human mind with the assistance of pen and paper. A human is able to observe and evaluate labels and determine if there are matching labels and apply an opinion based on the evaluation. The limitation is merely applying an abstract idea on generic computer system. See MPEP 2106.04(a)(2)(III)(c).
“in response to the identifying the congruent output label, selecting an explainability score for the live transaction from the set of precomputed explanations.” Under its broadest reasonable interpretation in light of the specification, this limitation encompasses the mental process of evaluating and observing data, which is an evaluation or observation that is practically capable of being performed in the human mind with the assistance of pen and paper. A human is able to evaluate data and identify if a label matches another observed label and selecting a value based on the evaluation. The limitation is merely applying an abstract idea on generic computer system. See MPEP 2106.04(a)(2)(III)(c).
Step 2A Prong 2 – Does the claim recite additional elements that integrate the judicial exception into a practical application?
This claim does not recite any additional limitations which integrate the abstract idea into a practical application.
Step 2B – Does the claim recite additional elements that amount to significantly more than the judicial exception?
Finally, the claim taken as a whole does not contain an inventive concept which provides significantly more than the abstract idea and thus the claim is subject-matter ineligible.
Claim 10
Step 1 – Is the claim to a process, machine, manufacture or composition of matter?
A process, as above.
Step 2A Prong 1 – Does the claim recite an abstract idea, law of nature, or natural phenomenon?
The claim recites the abstract ideas of the preceding claims from which it depends.
Step 2A Prong 2 – Does the claim recite additional elements that integrate the judicial exception into a practical application?
The claim recites the additional elements, “wherein the selected explainability score is a unified explainability score for the most aligned cluster.” amounts to generic computer components used as a tool to perform an existing process. Thus, the additional element amounts to no more than a recitation of the words "apply it" (or an equivalent) or are more than mere instructions to implement an abstract idea or other exception on a computer (see MPEP § 2106.05(f)).
Step 2B – Does the claim recite additional elements that amount to significantly more than the judicial exception?
Finally, the claim taken as a whole does not contain an inventive concept which provides significantly more than the abstract idea. The additional elements, “wherein the selected explainability score is a unified explainability score for the most aligned cluster.” amounts to generic computer components used as a tool to perform an existing process. Thus, the additional element amounts to no more than a recitation of the words "apply it" (or an equivalent) or are more than mere instructions to implement an abstract idea or other exception on a computer (see MPEP § 2106.05(f)).
Taken alone or in combination, the additional elements of the claim do not provide an inventive concept and thus the claim is subject-matter ineligible.
Claim 11
Step 1 – Is the claim to a process, machine, manufacture or composition of matter?
A process, as above.
Step 2A Prong 1 – Does the claim recite an abstract idea, law of nature, or natural phenomenon?
The claim recites the abstract ideas of the preceding claims from which it depends.
Step 2A Prong 2 – Does the claim recite additional elements that integrate the judicial exception into a practical application?
The claim recites the additional elements, “updating the most aligned cluster to include the input features of the live transaction.” amounts to generic computer components used as a tool to perform an existing process. Thus, the additional element amounts to no more than a recitation of the words "apply it" (or an equivalent) or are more than mere instructions to implement an abstract idea or other exception on a computer (see MPEP § 2106.05(f)).
Step 2B – Does the claim recite additional elements that amount to significantly more than the judicial exception?
Finally, the claim taken as a whole does not contain an inventive concept which provides significantly more than the abstract idea. The additional elements, “updating the most aligned cluster to include the input features of the live transaction.” amounts to generic computer components used as a tool to perform an existing process. Thus, the additional element amounts to no more than a recitation of the words "apply it" (or an equivalent) or are more than mere instructions to implement an abstract idea or other exception on a computer (see MPEP § 2106.05(f)).
Taken alone or in combination, the additional elements of the claim do not provide an inventive concept and thus the claim is subject-matter ineligible.
Claim 12
Step 1 – Is the claim to a process, machine, manufacture or composition of matter?
A process, as above.
Step 2A Prong 1 – Does the claim recite an abstract idea, law of nature, or natural phenomenon?
The claim recites the abstract ideas of the preceding claims from which it depends.
Step 2A Prong 2 – Does the claim recite additional elements that integrate the judicial exception into a practical application?
The claim recites the additional elements, “wherein the set of labelled transactions comprises units of work performed within the database management system against a database.” amounts to generic computer components used as a tool to perform an existing process. Thus, the additional element amounts to no more than a recitation of the words "apply it" (or an equivalent) or are more than mere instructions to implement an abstract idea or other exception on a computer (see MPEP § 2106.05(f)).
Step 2B – Does the claim recite additional elements that amount to significantly more than the judicial exception?
Finally, the claim taken as a whole does not contain an inventive concept which provides significantly more than the abstract idea. The additional elements, “wherein the set of labelled transactions comprises units of work performed within the database management system against a database.” amounts to generic computer components used as a tool to perform an existing process. Thus, the additional element amounts to no more than a recitation of the words "apply it" (or an equivalent) or are more than mere instructions to implement an abstract idea or other exception on a computer (see MPEP § 2106.05(f)).
Taken alone or in combination, the additional elements of the claim do not provide an inventive concept and thus the claim is subject-matter ineligible.
Claim 13 (Cancelled)
Claim 14
Step 1 – Is the claim to a process, machine, manufacture or composition of matter?
A process, as above.
Step 2A Prong 1 – Does the claim recite an abstract idea, law of nature, or natural phenomenon?
The claim recites the abstract ideas of the preceding claims from which it depends.
Step 2A Prong 2 – Does the claim recite additional elements that integrate the judicial exception into a practical application?
The claim recites the additional elements, “wherein the selecting the homogeneous region comprises selecting a branch level having a homogeneity score above an adjustable threshold homogeneity score.” amounts to generic computer components used as a tool to perform an existing process. Thus, the additional element amounts to no more than a recitation of the words "apply it" (or an equivalent) or are more than mere instructions to implement an abstract idea or other exception on a computer (see MPEP § 2106.05(f)).
Step 2B – Does the claim recite additional elements that amount to significantly more than the judicial exception?
Finally, the claim taken as a whole does not contain an inventive concept which provides significantly more than the abstract idea. The additional elements, “wherein the selecting the homogeneous region comprises selecting a branch level having a homogeneity score above an adjustable threshold homogeneity score.” amounts to generic computer components used as a tool to perform an existing process. Thus, the additional element amounts to no more than a recitation of the words "apply it" (or an equivalent) or are more than mere instructions to implement an abstract idea or other exception on a computer (see MPEP § 2106.05(f)).
Taken alone or in combination, the additional elements of the claim do not provide an inventive concept and thus the claim is subject-matter ineligible.
Claim 15
Step 1 – Is the claim to a process, machine, manufacture or composition of matter?
Claim 15 recites, “A system for outputting explanations for live transactions implemented by a database management system, comprising: a memory; and a processor communicatively coupled to the memory, wherein the processor is configured to perform a method comprising:” therefore it is directed to the statutory category of a machine.
Step 2A Prong 1 – Does the claim recite an abstract idea, law of nature, or natural phenomenon?
The claim recites, inter alia:
“hierarchically clustering an input space of the labeled transactions based on the input features to form input space clusters of the labeled transactions divided into branch levels;” Under its broadest reasonable interpretation in light of the specification, this limitation encompasses mathematical concept of utilizing a mathematical formula to perform calculations. Hierarchical clustering is a mathematical process that can be completed by a human with pen and paper or using a computer as a tool. This claim discloses a math operation and therefore is ineligible.
“identifying a homogenous region of the input space, wherein the identifying comprises:” Under its broadest reasonable interpretation in light of the specification, this limitation encompasses the mental process of evaluating and observing data, which is an evaluation or observation that is practically capable of being performed in the human mind with the assistance of pen and paper. A human is able to evaluate data and identify a cluster. The limitation is merely applying an abstract idea on generic computer system. See MPEP 2106.04(a)(2)(III)(c).
“scoring homogeneity of the input space clusters based on a number of matching output labels from the output labels corresponding to the labeled transactions in each input space cluster;” Under its broadest reasonable interpretation in light of the specification, this limitation encompasses the mental process of evaluating and observing data, which is an evaluation or observation that is practically capable of being performed in the human mind with the assistance of pen and paper. A human is able to evaluate clusters and provide a value as an opinion or judgment. The limitation is merely applying an abstract idea on generic computer system. See MPEP 2106.04(a)(2)(III)(c).
“scoring homogeneity of the branch levels based on the homogeneity scoring of the input space clusters; and” Under its broadest reasonable interpretation in light of the specification, this limitation encompasses the mental process of evaluating and observing data, which is an evaluation or observation that is practically capable of being performed in the human mind with the assistance of pen and paper. A human is able to evaluate data and apply values to the data based on judgement. The limitation is merely applying an abstract idea on generic computer system. See MPEP 2106.04(a)(2)(III)(c).
“selecting the homogeneous region based on the homogeneity scoring of the branch levels;” Under its broadest reasonable interpretation in light of the specification, this limitation encompasses the mental process of evaluating and observing data, which is an evaluation or observation that is practically capable of being performed in the human mind with the assistance of pen and paper. A human is able to evaluate data and apply a value based on judgement. The limitation is merely applying an abstract idea on generic computer system. See MPEP 2106.04(a)(2)(III)(c).
“selecting, by the processor using the XAI module, from the set of precomputed explanations, an explainability score for a live transaction.” Under its broadest reasonable interpretation in light of the specification, this limitation encompasses the mental process of evaluating and observing data, which is an evaluation or observation that is practically capable of being performed in the human mind with the assistance of pen and paper. A human is able to evaluate data and make a decision or opinion based on the evaluation. The limitation is merely applying an abstract idea on generic computer system. See MPEP 2106.04(a)(2)(III)(c).
Step 2A Prong 2 – Does the claim recite additional elements that integrate the judicial exception into a practical application?
The claim recites the additional elements, “generating, by the processor, an explainable artificial intelligence (XAI) module, wherein the generating comprises:” amounts to generic computer components used as a tool to perform an existing process. Thus, the additional element amounts to no more than a recitation of the words "apply it" (or an equivalent) or are more than mere instructions to implement an abstract idea or other exception on a computer (see MPEP § 2106.05(f)).
“obtaining a set of labeled transactions completed by the database management system, the labeled transactions comprising input features and corresponding output labels generated by a machine learning (ML) model;” is an insignificant extra-solution activity required for any uses of the mental processes (see MPEP § 2106.05(g)) As such, the claim is ineligible.
“obtaining, by an explainability model, explainability scores for labeled transactions in the homogeneous region;” amounts to generic computer components used as a tool to perform an existing process. Thus, the additional element amounts to no more than a recitation of the words "apply it" (or an equivalent) or are more than mere instructions to implement an abstract idea or other exception on a computer (see MPEP § 2106.05(f)).
“generating a unified explainability score for each input space cluster based on the explainability scores; and” amounts to generic computer components used as a tool to perform an existing process. Thus, the additional element amounts to no more than a recitation of the words "apply it" (or an equivalent) or are more than mere instructions to implement an abstract idea or other exception on a computer (see MPEP § 2106.05(f)).
“storing each unified explainability score in a set of precomputed explanations; and” is an insignificant extra-solution activity required for any uses of the mental processes (see MPEP § 2106.05(g)) As such, the claim is ineligible.
Step 2B – Does the claim recite additional elements that amount to significantly more than the judicial exception?
Finally, the claim taken as a whole does not contain an inventive concept which provides significantly more than the abstract idea. The additional elements, “generating, by the processor, an explainable artificial intelligence (XAI) module, wherein the generating comprises:” amounts to generic computer components used as a tool to perform an existing process. Thus, the additional element amounts to no more than a recitation of the words "apply it" (or an equivalent) or are more than mere instructions to implement an abstract idea or other exception on a computer (see MPEP § 2106.05(f)).
“obtaining a set of labeled transactions completed by the database management system, the labeled transactions comprising input features and corresponding output labels generated by a machine learning (ML) model;” is an insignificant extra-solution activity required for any uses of abstract ideas (see MPEP § 2106.05(g)), and is a well-understood, routine, conventional activity (see MPEP § 2106.05(d)(II)(i); “Receiving or transmitting data over a network, e.g., using the Internet to gather data”.
“obtaining, by an explainability model, explainability scores for labeled transactions in the homogeneous region;” amounts to generic computer components used as a tool to perform an existing process. Thus, the additional element amounts to no more than a recitation of the words "apply it" (or an equivalent) or are more than mere instructions to implement an abstract idea or other exception on a computer (see MPEP § 2106.05(f)).
“generating a unified explainability score for each input space cluster based on the explainability scores; and” amounts to generic computer components used as a tool to perform an existing process. Thus, the additional element amounts to no more than a recitation of the words "apply it" (or an equivalent) or are more than mere instructions to implement an abstract idea or other exception on a computer (see MPEP § 2106.05(f)).
“storing each unified explainability score in a set of precomputed explanations; and” is an insignificant extra-solution activity required for any uses of abstract ideas (see MPEP § 2106.05(g)), and is a well-understood, routine, conventional activity (see MPEP § 2106.05(d)(iv); “Storing and retrieving information in memory”.
Taken alone or in combination, the additional elements of the claim do not provide an inventive concept and thus the claim is subject-matter ineligible.
Claim 16
Step 1 – Is the claim to a process, machine, manufacture or composition of matter?
A machine, as above.
Step 2A Prong 1 – Does the claim recite an abstract idea, law of nature, or natural phenomenon?
The claim recites the abstract ideas of the preceding claims from which it depends.
Step 2A Prong 2 – Does the claim recite additional elements that integrate the judicial exception into a practical application?
The claim recites the additional elements, “receiving the live transaction from a user device; and” is an insignificant extra-solution activity required for any uses of the mental processes (see MPEP § 2106.05(g)) As such, the claim is ineligible.
“extracting, by the XAI module, input features from the live transaction.” amounts to generic computer components used as a tool to perform an existing process. Thus, the additional element amounts to no more than a recitation of the words "apply it" (or an equivalent) or are more than mere instructions to implement an abstract idea or other exception on a computer (see MPEP § 2106.05(f)).
Step 2B – Does the claim recite additional elements that amount to significantly more than the judicial exception?
Finally, the claim taken as a whole does not contain an inventive concept which provides significantly more than the abstract idea. The additional elements, “receiving the live transaction from a user device; and” is an insignificant extra-solution activity required for any uses of abstract ideas (see MPEP § 2106.05(g)), and is a well-understood, routine, conventional activity (see MPEP § 2106.05(d)(II)(i); “Receiving or transmitting data over a network, e.g., using the Internet to gather data”.
“extracting, by the XAI module, input features from the live transaction.” amounts to generic computer components used as a tool to perform an existing process. Thus, the additional element amounts to no more than a recitation of the words "apply it" (or an equivalent) or are more than mere instructions to implement an abstract idea or other exception on a computer (see MPEP § 2106.05(f)).
Taken alone or in combination, the additional elements of the claim do not provide an inventive concept and thus the claim is subject-matter ineligible.
Claim 17
Step 1 – Is the claim to a process, machine, manufacture or composition of matter?
A machine, as above.
Step 2A Prong 1 – Does the claim recite an abstract idea, law of nature, or natural phenomenon?
The claim recites the abstract ideas of the preceding claims from which it depends.
Step 2A Prong 2 – Does the claim recite additional elements that integrate the judicial exception into a practical application?
The claim recites the additional elements, “wherein the explainability score is selected based on the input features of the live transaction.” amounts to generic computer components used as a tool to perform an existing process. Thus, the additional element amounts to no more than a recitation of the words "apply it" (or an equivalent) or are more than mere instructions to implement an abstract idea or other exception on a computer (see MPEP § 2106.05(f)).
Step 2B – Does the claim recite additional elements that amount to significantly more than the judicial exception?
Finally, the claim taken as a whole does not contain an inventive concept which provides significantly more than the abstract idea. The additional elements, “wherein the explainability score is selected based on the input features of the live transaction.” amounts to generic computer components used as a tool to perform an existing process. Thus, the additional element amounts to no more than a recitation of the words "apply it" (or an equivalent) or are more than mere instructions to implement an abstract idea or other exception on a computer (see MPEP § 2106.05(f)).
Taken alone or in combination, the additional elements of the claim do not provide an inventive concept and thus the claim is subject-matter ineligible.
Claim 18
Step 1 – Is the claim to a process, machine, manufacture or composition of matter?
Claim 18 recites, “A computer program product, the computer program product comprising a computer readable storage medium having program instructions embodied therewith, the program instructions executable by a processor to cause a device to perform a method of outputting explanations for live transactions implemented by a database management system, the method comprising:” therefore it is directed to the statutory category of a machine.
Step 2A Prong 1 – Does the claim recite an abstract idea, law of nature, or natural phenomenon?
The claim recites, inter alia:
“hierarchically clustering an input space of the labeled transactions based on the input features to form input space clusters of the labeled transactions divided into branch levels;” Under its broadest reasonable interpretation in light of the specification, this limitation encompasses mathematical concept of utilizing a mathematical formula to perform calculations. Hierarchical clustering is a mathematical process that can be completed by a human with pen and paper or using a computer as a tool. This claim discloses a math operation and therefore is ineligible.
“identifying a homogenous region of the input space, wherein the identifying comprises:” Under its broadest reasonable interpretation in light of the specification, this limitation encompasses the mental process of evaluating and observing data, which is an evaluation or observation that is practically capable of being performed in the human mind with the assistance of pen and paper. A human is able to evaluate data and identify a cluster. The limitation is merely applying an abstract idea on generic computer system. See MPEP 2106.04(a)(2)(III)(c).
“scoring homogeneity of the input space clusters based on a number of matching output labels from the corresponding output labels corresponding to the labeled transactions in each input space cluster;” Under its broadest reasonable interpretation in light of the specification, this limitation encompasses the mental process of evaluating and observing data, which is an evaluation or observation that is practically capable of being performed in the human mind with the assistance of pen and paper. A human is able to evaluate clusters and provide a value as an opinion or judgment. The limitation is merely applying an abstract idea on generic computer system. See MPEP 2106.04(a)(2)(III)(c).
“scoring homogeneity of the branch levels based on the homogeneity scoring of the input space clusters; and” Under its broadest reasonable interpretation in light of the specification, this limitation encompasses the mental process of evaluating and observing data, which is an evaluation or observation that is practically capable of being performed in the human mind with the assistance of pen and paper. A human is able to evaluate data and apply values to the data based on judgement. The limitation is merely applying an abstract idea on generic computer system. See MPEP 2106.04(a)(2)(III)(c).
“selecting the homogeneous region based on the homogeneity scoring of the branch levels;” Under its broadest reasonable interpretation in light of the specification, this limitation encompasses the mental process of evaluating and observing data, which is an evaluation or observation that is practically capable of being performed in the human mind with the assistance of pen and paper. A human is able to evaluate data and apply a value based on judgement. The limitation is merely applying an abstract idea on generic computer system. See MPEP 2106.04(a)(2)(III)(c).
“selecting, by the processor using the XAI module, from the set of precomputed explanations, an explainability score for a live transaction.” Under its broadest reasonable interpretation in light of the specification, this limitation encompasses the mental process of evaluating and observing data, which is an evaluation or observation that is practically capable of being performed in the human mind with the assistance of pen and paper. A human is able to evaluate data and make a decision or opinion based on the evaluation. The limitation is merely applying an abstract idea on generic computer system. See MPEP 2106.04(a)(2)(III)(c).
Step 2A Prong 2 – Does the claim recite additional elements that integrate the judicial exception into a practical application?
The claim recites the additional elements, “generating, by the processor, an explainable artificial intelligence (XAI) module, wherein the generating comprises:” amounts to generic computer components used as a tool to perform an existing process. Thus, the additional element amounts to no more than a recitation of the words "apply it" (or an equivalent) or are more than mere instructions to implement an abstract idea or other exception on a computer (see MPEP § 2106.05(f)).
“obtaining, a set of labeled transactions completed by the database management system, the labeled transactions comprising input features and corresponding output labels generated by a machine learning (ML) model;” is an insignificant extra-solution activity required for any uses of the mental processes (see MPEP § 2106.05(g)) As such, the claim is ineligible.
“obtaining, by an explainability model, explainability scores for labeled transactions in the homogeneous region;” amounts to generic computer components used as a tool to perform an existing process. Thus, the additional element amounts to no more than a recitation of the words "apply it" (or an equivalent) or are more than mere instructions to implement an abstract idea or other exception on a computer (see MPEP § 2106.05(f)).
“generating a unified explainability score for each input space cluster in the homogeneous region based on the explainability scores; and” amounts to generic computer components used as a tool to perform an existing process. Thus, the additional element amounts to no more than a recitation of the words "apply it" (or an equivalent) or are more than mere instructions to implement an abstract idea or other exception on a computer (see MPEP § 2106.05(f)).
“storing each unified explainability score in a set of precomputed explanations; and” is an insignificant extra-solution activity required for any uses of the mental processes (see MPEP § 2106.05(g)) As such, the claim is ineligible.
Step 2B – Does the claim recite additional elements that amount to significantly more than the judicial exception?
Finally, the claim taken as a whole does not contain an inventive concept which provides significantly more than the abstract idea. The additional elements, “generating, by the processor, an explainable artificial intelligence (XAI) module, wherein the generating comprises:” amounts to generic computer components used as a tool to perform an existing process. Thus, the additional element amounts to no more than a recitation of the words "apply it" (or an equivalent) or are more than mere instructions to implement an abstract idea or other exception on a computer (see MPEP § 2106.05(f)).
“obtaining, a set of labeled transactions completed by the database management system, the labeled transactions comprising input features and corresponding output labels generated by a machine learning (ML) model;” is an insignificant extra-solution activity required for any uses of abstract ideas (see MPEP § 2106.05(g)), and is a well-understood, routine, conventional activity (see MPEP § 2106.05(d)(II)(i); “Receiving or transmitting data over a network, e.g., using the Internet to gather data”.
“obtaining, by an explainability model, explainability scores for labeled transactions in the homogeneous region;” amounts to generic computer components used as a tool to perform an existing process. Thus, the additional element amounts to no more than a recitation of the words "apply it" (or an equivalent) or are more than mere instructions to implement an abstract idea or other exception on a computer (see MPEP § 2106.05(f)).
“generating a unified explainability score for each input space cluster in the homogeneous region based on the explainability scores; and” amounts to generic computer components used as a tool to perform an existing process. Thus, the additional element amounts to no more than a recitation of the words "apply it" (or an equivalent) or are more than mere instructions to implement an abstract idea or other exception on a computer (see MPEP § 2106.05(f)).
“storing each unified explainability score in a set of precomputed explanations; and” is an insignificant extra-solution activity required for any uses of abstract ideas (see MPEP § 2106.05(g)), and is a well-understood, routine, conventional activity (see MPEP § 2106.05(d)(iv); “Storing and retrieving information in memory”.
Taken alone or in combination, the additional elements of the claim do not provide an inventive concept and thus the claim is subject-matter ineligible.
Claim 19
Step 1 – Is the claim to a process, machine, manufacture or composition of matter?
A machine, as above.
Step 2A Prong 1 – Does the claim recite an abstract idea, law of nature, or natural phenomenon?
The claim recites the abstract ideas of the preceding claims from which it depends.
Step 2A Prong 2 – Does the claim recite additional elements that integrate the judicial exception into a practical application?
The claim recites the additional elements, “receiving the live transaction from a user device; and” is an insignificant extra-solution activity required for any uses of the mental processes (see MPEP § 2106.05(g)) As such, the claim is ineligible.
“extracting, by the XAI module, input features from the live transaction.” amounts to generic computer components used as a tool to perform an existing process. Thus, the additional element amounts to no more than a recitation of the words "apply it" (or an equivalent) or are more than mere instructions to implement an abstract idea or other exception on a computer (see MPEP § 2106.05(f)).
Step 2B – Does the claim recite additional elements that amount to significantly more than the judicial exception?
Finally, the claim taken as a whole does not contain an inventive concept which provides significantly more than the abstract idea. The additional elements, “receiving the live transaction from a user device; and” is an insignificant extra-solution activity required for any uses of abstract ideas (see MPEP § 2106.05(g)), and is a well-understood, routine, conventional activity (see MPEP § 2106.05(d)(II)(i); “Receiving or transmitting data over a network, e.g., using the Internet to gather data”.
“extracting, by the XAI module, input features from the live transaction.” amounts to generic computer components used as a tool to perform an existing process. Thus, the additional element amounts to no more than a recitation of the words "apply it" (or an equivalent) or are more than mere instructions to implement an abstract idea or other exception on a computer (see MPEP § 2106.05(f)).
Taken alone or in combination, the additional elements of the claim do not provide an inventive concept and thus the claim is subject-matter ineligible.
Claim 20
Step 1 – Is the claim to a process, machine, manufacture or composition of matter?
A machine, as above.
Step 2A Prong 1 – Does the claim recite an abstract idea, law of nature, or natural phenomenon?
The claim recites the abstract ideas of the preceding claims from which it depends.
Step 2A Prong 2 – Does the claim recite additional elements that integrate the judicial exception into a practical application?
The claim recites the additional elements, “wherein the explainability score is selected based on the input features of the live transaction.” amounts to generic computer components used as a tool to perform an existing process. Thus, the additional element amounts to no more than a recitation of the words "apply it" (or an equivalent) or are more than mere instructions to implement an abstract idea or other exception on a computer (see MPEP § 2106.05(f)).
Step 2B – Does the claim recite additional elements that amount to significantly more than the judicial exception?
Finally, the claim taken as a whole does not contain an inventive concept which provides significantly more than the abstract idea. The additional elements, “wherein the explainability score is selected based on the input features of the live transaction.” amounts to generic computer components used as a tool to perform an existing process. Thus, the additional element amounts to no more than a recitation of the words "apply it" (or an equivalent) or are more than mere instructions to implement an abstract idea or other exception on a computer (see MPEP § 2106.05(f)).
Taken alone or in combination, the additional elements of the claim do not provide an inventive concept and thus the claim is subject-matter ineligible.
Claim 21
Step 1 – Is the claim to a process, machine, manufacture or composition of matter?
A process, as above.
Step 2A Prong 1 – Does the claim recite an abstract idea, law of nature, or natural phenomenon?
The claim recites, inter alia:
“identifying a most aligned cluster from the homogeneous region based on input features of the live transaction; and” Under its broadest reasonable interpretation in light of the specification, this limitation encompasses the mental process of evaluating and observing data, which is an evaluation or observation that is practically capable of being performed in the human mind with the assistance of pen and paper. A human is able to evaluate a set of clusters and judge which cluster is the most aligned. The limitation is merely applying an abstract idea on generic computer system. See MPEP 2106.04(a)(2)(III)(c).
“determining that the most aligned cluster satisfies a closeness criterion;” Under its broadest reasonable interpretation in light of the specification, this limitation encompasses the mental process of evaluating and observing data, which is an evaluation or observation that is practically capable of being performed in the human mind with the assistance of pen and paper. A human is able to evaluate a cluster to determine it meet criteria and provide an opinion. The limitation is merely applying an abstract idea on generic computer system. See MPEP 2106.04(a)(2)(III)(c).
“in response to determining that the most aligned cluster satisfies the closeness criterion, comparing output labels of the most aligned cluster with an output label of the live transaction.” Under its broadest reasonable interpretation in light of the specification, this limitation encompasses the mental process of evaluating and observing data, which is an evaluation or observation that is practically capable of being performed in the human mind with the assistance of pen and paper. A human is able to evaluate data and compare labels and provide and opinion or judgement based on the comparison. The limitation is merely applying an abstract idea on generic computer system. See MPEP 2106.04(a)(2)(III)(c).
Step 2A Prong 2 – Does the claim recite additional elements that integrate the judicial exception into a practical application?
This claim does not recite any additional limitations which integrate the abstract idea into a practical application.
Step 2B – Does the claim recite additional elements that amount to significantly more than the judicial exception?
Finally, the claim taken as a whole does not contain an inventive concept which provides significantly more than the abstract idea and thus the claim is subject-matter ineligible.
Claim 22
Step 1 – Is the claim to a process, machine, manufacture or composition of matter?
A process, as above.
Step 2A Prong 1 – Does the claim recite an abstract idea, law of nature, or natural phenomenon?
The claim recites the abstract ideas of the preceding claims from which it depends.
Step 2A Prong 2 – Does the claim recite additional elements that integrate the judicial exception into a practical application?
The claim recites the additional elements, “wherein the generating the XAI module further comprises storing the explainability scores for the labeled transactions in the set of precomputed explanations.” is an insignificant extra-solution activity required for any uses of the mental processes (see MPEP § 2106.05(g)) As such, the claim is ineligible.
Step 2B – Does the claim recite additional elements that amount to significantly more than the judicial exception?
Finally, the claim taken as a whole does not contain an inventive concept which provides significantly more than the abstract idea. The additional elements, “wherein the generating the XAI module further comprises storing the explainability scores for the labeled transactions in the set of precomputed explanations.” is an insignificant extra-solution activity required for any uses of abstract ideas (see MPEP § 2106.05(g)), and is a well-understood, routine, conventional activity (see MPEP § 2106.05(d)(iv); “Storing and retrieving information in memory”.
Taken alone or in combination, the additional elements of the claim do not provide an inventive concept and thus the claim is subject-matter ineligible.
Claim Rejections - 35 USC § 103
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
Claims 1, 2, 4-12, and 14-22 are rejected under 35 U.S.C. 103 as being unpatentable over Dalli et al, (Dalli et al, “AUTOMATIC XAI (AUTOXAI) WITH EVOLUTIONARY NAS TECHNIQUES AND MODEL DISCOVERY AND REFINEMENT”, US 20220398460 A1, Filed Jun. 9th, 2022, hereinafter “Dalli”) in view of Zhao et al, (Zhao et al, “Evaluation of Hierarchical Clustering Algorithms for Document Datasets”, 2022, hereinafter “Zhao”).
Regarding claim 1, Dalli discloses, “A method of outputting explanations for live transactions implemented by a database management system, the database management system comprising a processor communicatively coupled to a memory, and the method comprising:” (Detailed Description, pp. 37, [0435]; “In another exemplary embodiment, AutoXAI may be integrated with a workflow system with a bi-directional exchange of information between the AutoXAI system and the workflow system, including both processing data and event data. Symbolic information within the AutoXAI system may be made accessible to the workflow system, which can then take automated action accordingly. Conversely, workflow system information and workflow transition states or operational states may be made accessible to the AutoXAI system for control and configuration purposes. It is further contemplated that the AutoXAI and workflow combination may be further integrated within a Robotic Process Automation (RPA) system, Decision Support System (DSS) or a Data Lake system.” The XAI module in this application is modular and able to be implemented in different settings. As stated above, this system would be able to integrate into a workflow system of a data lake and provide explanations based on information in the data lake or analyze data generated by a user.)
“generating, by the processor, an explainable artificial intelligence (XAI) module, wherein the generating comprises: obtaining a set of labeled transactions completed by the database management system, the labeled transactions comprising input features and corresponding output labels generated by a machine learning (ML) model;” (Detailed Description, pp. 3, [0045]; “An exemplary embodiment may implement an explainable neural network (which may be considered a form of XAI). Referring now to exemplary FIG. 12, FIG. 12 may illustrate a schematic diagram of an exemplary high level XNN architecture.” The model in this application is able to generate a XAI modules to be used in a multi-modal system. Fig. 12 is an example of a XNN module.) and (Detailed Description, pp. 4, [0062]; “Hierarchical clustering techniques or other logically equivalent methods can be used for identifying suitable partitions, such as an XAI model induction method which may input a set of training data to a black-box system and analyze the corresponding output using a partitioning function.” The XAI module in this application is able to evaluate input data and output data from another ML model. Further the system is able to cluster data using hierarchical techniques.)
“hierarchically clustering an input space of the labeled transactions based on the input features to form input space clusters of the labeled transactions divided into branch levels;” (Detailed Description, pp. 4, [0064]; “The input to the partitioning method can be either the input features directly for low-dimensional data (i.e., tabular) or data which has been pre-processed (for example, from a convolutional network). Features which have been transformed using a convolutional process may typically represent a higher-level of abstraction such as an edge, a stroke, or a pattern.” The system in this application is able to cluster, or partition, data using hierarchal clustering. This system will cluster data based on the input data and model output data.) and (Detailed Description, pp. 4, [0063]; “In an exemplary embodiment that may use a hierarchical clustering method for partitioning, a variety of appropriate methods may be used for a practical implementation including, but not limited to, agglomerative clustering, divisive clustering, relocation partitioning, probabilistic clustering, k-medoid methods, k-means methods, fuzzy clustering, density based clustering, grid based methods, gradient descent based methods, evolutionary methods, region splitting, region growing, sub-space clustering, projection methods, co-clustering methods and lazy clustering methods.” The proposed system is able to perform many different forms of hierarchical clustering. As disclosed, agglomerative clustering may be used which utilizes a tree structure containing branches and nodes/leaf.)
“obtaining, in response to the identifying and by an explainability model, explainability scores for labeled transactions in each input space cluster in the homogeneous region;” (Detailed Description, pp. 14, [0189]; “A partition is a region in the data, which may be disjointing or overlapping. A rule may be a linear or nonlinear equation which may include coefficients with their respective dimension, and the result may represent both the answer to the problem and the explanation coefficients which may be used to generate domain specific explanations that are both machine and human readable. A rule may further represent a justification that explains how the explanation itself was produced. An exemplary embodiment applies an element of human readability to the encoded knowledge, data and rules which are otherwise too complex for an ordinary person to reproduce or comprehend without any automated process.” The model in this application is able to establish rules or a set of rules pertaining to the final explanation. This process can be performed on different partitions or clusters generated earlier. As stated above, a rule can be derived from a mathematical transformation or equation. Further, each of the partitions or spaces containing similar features can include assigned values which is used to generate explanations.)
“generating a unified explainability score for each input space cluster in the homogeneous region based on the explainability scores; and” (Detailed Description, pp. 14, [0186-0187]; “[0186] Still referring to exemplary FIG. 16, the method may continue by aggregating the data point predictions or classifications into hierarchal partitions 208. Rule conditions may be obtained from the hierarchal partitions. An external function defined by Partition(X) may identify the partitions. Partition(X) may be a function configured to partition similar data and may be used to create rules. The partitioning function may include a clustering algorithm such as k-means, entropy, or a mutual information-based method. [0187] The hierarchical partitions may organize the output data points in a variety of ways. In an exemplary embodiment, the data points may be aggregated such that each partition represents a rule or a set of rules. The hierarchical partitions may then be modeled using mathematical transformations and linear models.” As stated above, each partition or cluster can have rules applied. These rules can perform mathematical functions and are able to generate values for a partition. The partition or clusters are designed to group similar features to space or region using rules or a set of rules. This information will be used to generate an explanation.)
“storing each unified explainability score in a set of precomputed explanations; and” (Detailed Description, pp. 3, [0046]; “The layers may be analyzed by the selection and ranking layer 128 that may multiply the switch output by the value output, producing a ranked or scored output 130. The explanations and answers may be concurrently calculated by the XNN by the conditional network and the prediction network. The selection and ranking layer 128 may ensure that the answers and explanations are correctly matched, ranked and scored appropriately before being sent to the output 130.” The model in this application is part of a multi-model system. As stated above, some components or models run concurrently. The actions of some models may be completed and stored which are retrieved and evaluated at a later point to output an explanation.)
“selecting, by the processor using the XAI module, and from the set of precomputed explanations, an explainability score for a live transaction.” (Detailed Description, pp. 7, [0099]; “The end result 914 of the explanation process may include either an explanation and/or its interpretation, which may be consumed by a human user, another application, another system component forming part of a larger embodiment, or some other automated system.” The model in this application is able to output an explanation based on input information.) and (Detailed Description, pp. 8, [0104-0105]; “The explanation scaffolding 9101 can also store audit data in appropriate components such as the actions 1538 or the evaluation 1535 components, or via an independent audit system plugged in to the third-party data 1539 extension component. [0105] The interpretation scaffolding 9111, illustrated in FIG. 15C, can also store audit data in appropriate components such as the interactive context 1544 or the protocol context 1548 components. The interpretation scaffolding 9111 may be structured into three main components, the explanation and interpretation scenario component 9112, the framing, protocol, and contextual component 1540, and the interpretation model component 1550.” This model also uses stored values from previous explanations to produce current explanations. This teaches that the output of this model is generated based on or uses prior computer values and outcomes.)
Dalli fails to explicitly disclose, “identifying a homogenous region of the input space, wherein the identifying comprises: scoring homogeneity of the input space clusters based on a number of matching output labels from the output labels corresponding to the labeled transactions in each input space cluster;”, “scoring homogeneity of the branch levels based on the homogeneity scoring of the input space clusters; and” and “selecting the homogeneous region based on the homogeneity scoring of the branch levels;”.
However, Zhao discloses, “identifying a homogenous region of the input space, wherein the identifying comprises: scoring homogeneity of the input space clusters based on a number of matching output labels from the output labels corresponding to the labeled transactions in each input space cluster;” (Internal Criterion Functions, pp. 517; “The first internal criterion function maximizes the sum of the average pairwise similarities between the documents assigned to each cluster, weighted according to the size of each cluster. Specifically, if we use the cosine function to measure the similarity between documents, then we want the clustering solution to minimize the following criterion function: [See Equation (3)].” Zhao discloses different methods of clustering and gives a description to the different scoring criterion used in clustering documents using hierarchical clustering. This teaches the evaluation of cluster to determine similarity in the cluster. This is used to generate a homogenous cluster.)
“scoring homogeneity of the branch levels based on the homogeneity scoring of the input space clusters; and” (Internal Criterion Functions, pp. 517; “The second criterion function is used by the popular vector-space variant of the K-means algorithm [6, 18, 7, 26, 15]. In this algorithm each cluster is represented by its centroid vector and the goal is to find the clustering solution that maximizes the similarity between each document and the centroid of the cluster that is assigned to. Specifically, if we use the cosine function to measure the similarity between a document and a centroid, then the criterion function maximizes the following: [See Equation (4)].” Zhao discloses further criterion for clustering which involve evaluating a clusters content to the center of the cluster. This teaches an evaluation of the different branch levels or clusters.)
“selecting the homogeneous region based on the homogeneity scoring of the branch levels;” (External Criterion Functions, pp. 517; “As we can see from Equation 5, even-though our initial motivation was to define an external criterion function, because we used the cosine function to measure the separation between the cluster and the entire collection, the criterion function does take into account the within-cluster similarity of the documents (due to the ||Dr|| term). Thus, ℇ1 is actually a hybrid criterion function that combines both external and internal characteristics of the clusters.” Zhao discloses a process of evaluating a cluster or branch to the whole model. This is disclosed as an external criterion. This process including selecting a cluster and evaluating it.)
It would have been obvious to one of ordinary skill in the art before the effective filing date of the present application to combine Dalli and Zhao. Dalli teaches a system that is able to evaluate input data and determine which XAI model could be used to explain the output of various machine learning models. Zhao teaches the use of Hierarchical Clustering to cluster documents based on similar elements of the documents and provides a comparison of different clustering algorithms. One of ordinary skill would have motivation to combine a system which is able to provide a XAI model which can work with many different kinds of AI models and system that perform Hierarchical Clustering to evaluate data, “A number of observations can be made by analyzing the results in Table 5. First, for all the datasets except tr41, the constrained agglomerative methods improved the hierarchical solutions obtained by agglomerative methods alone, no matter what partitional clustering algorithm is used to obtain intermediate clusters. The improvement can be achieved even with small number of intermediate clusters. Second, for many cases, the constrained agglomerative methods performed even better than the corresponding partitional methods. Finally, the partitional clustering methods that improved the agglomerative hierarchical results the most are the same partitional clustering methods that performed the best in terms of generating the whole hierarchical trees.” (Zhao, Constrained Agglomerative Trees, pp. 522).
Regarding claim 2, Dalli discloses, “receiving a second live transaction from a user device.” (Detailed Description, pp. 37, [0434]; “In another exemplary embodiment, an AutoXAI may be integrated with a Robotic Process Automation (RPA) system with a bi-directional exchange of information between the AutoXAI system and the RPA system. Symbolic information within the AutoXAI system may be made accessible to the RPA system, which can then take automated action accordingly.” The process disclosed in Dalli can be applied to various different models. This can include, as stated above, a robotic system. This system is able to work online and live able to take and evaluate transactions provided in updates or communication.)
Regarding claim 4, Dalli discloses, “wherein the explainability score is selected based on input features of the live transaction.” (Detailed Description, pp. 36, [0427]; “In parts of the flight envelope, some input may lead to dangerous instability or even irrecoverable situations, such as "deep stall" or "flat spin", respectively. An AI system may be able to analyze these situations in order to prevent them from happening. However, an exemplary XAI system may be able to explain which input configurations are causing dangerous flight conditions and offer suggested changes as part of the explanation, so that the pilot may quickly adjust the control stick input to recover good flight characteristics or, if deemed necessary, retain the current input.” The XAI system in this application is designed to be modular. As stated above this system can run in real time and produce explanations based on regular inputs or transactions from an active system.)
Regarding claim 5, Zhao discloses, “selecting, by the XAI module, a most aligned cluster from the homogeneous region based on input features of the second live transaction; and” (Clustering, pp. 8; "The refinement strategy that we used consists of a number of iterations. During each iteration, the documents are visited in a random order. For each document, di, we compute the change in the value of the criterion function obtained by moving di to one of the other k - 1 clusters. If there exist some moves that lead to an improvement in the overall value of the criterion function, then di is moved to the cluster that leads to the highest improvement." During training each of the pairs are selected. This can include the most aligned and least aligned cluster to look for improvements. This process is executed iteratively, evaluating one to many documents, or transactions, in different clusters.)
“determining, by the XAI module, whether the most aligned cluster satisfies a closeness criterion.” (Criterion Function Optimization, pp. 518; "If there exist some moves that lead to an improvement in the overall value of the criterion function, then di is moved to the cluster that leads to the highest improvement. If no such cluster exists, di remains in the cluster that it already belongs to. The refinement phase ends, as soon as we perform an iteration in which no documents moved between clusters." Each of the clusters are evaluated. While being evaluated if there is another cluster with a better match di is moved to another cluster. This teaches that the most aligned and the least aligned clusters are evaluated on certain criteria for closeness.)
Regarding claim 6, Zhao discloses, “in response to determining that the most aligned cluster does not satisfy the closeness criterion, generating a new explainability score for the second live transaction.” (Criterion Function Optimization, pp. 518; "The refinement strategy that we used consists of a number of iterations. During each iteration, the documents are visited in a random order. For each document, di, we compute the change in the value of the criterion function obtained by moving di to one of the other k -1 clusters. If there exist some moves that lead to an improvement in the overall value of the criterion function, then di is moved to the cluster that leads to the highest improvement." Each of the data points are evaluated during the training process. There is a threshold that is met and a score is generated. If there is a closer match to di then a new score is assigned and di will be moved to another cluster. This teaches that a new score is assigned based on a threshold. This process is executed iteratively, evaluating one to many documents, or transactions, in different clusters.)
Regarding claim 7, Zhao discloses, “in response to determining that the most aligned cluster satisfies the closeness criterion, comparing output labels of the most aligned cluster with output labels of the second live transaction.” (Preliminaries, pp. 516; "The various clustering algorithms that are described in this paper use the vector-space model [24] to represent each document. In this model, each document d is considered to be a vector in the term-space." The model in this article is used to cluster documents. As stated above when clustering the model will determine a criterion based on the data type.) and (Criterion Function Optimization, pp. 518; “Initially, a random pair of documents is selected from the collection to act as the seeds of the two clusters. Then, for each document, its similarity to these two seeds is computed, and it is assigned to the cluster corresponding to its most similar seed. This forms the initial two-way clustering. This clustering is then repeatedly refined so that it optimizes the desired clustering criterion function.” This system will start with two documents and compare them to each other. If the documents are similar, the values are mapped to each other in the cluster. As each document is added the documents of the nodes are compared and evaluated for closeness.)
Regarding claim 8, Zhao discloses, “determining, based on the comparing, that the output labels do not include a congruent output label; and” (Preliminaries, pp. 516; "The cosine formula can be simplified to
cos
d
i
,
d
j
=
d
i
t
d
j
, when the document vectors are of unit length. This measure becomes one if the documents are identical, and zero if there is nothing in common between them (i.e., the vectors are orthogonal to each other)." This system will evaluate the document for congruency. If the documents are the same or identical they will receive a specified value denoting so.)
“in response to the determining that the output labels do not include a congruent output label, generating a new explainability score for the second live transaction.” (Criterion Function Optimization, pp. 518; "The refinement strategy that we used consists of a number of iterations. During each iteration, the documents are visited in a random order. For each document, di, we compute the change in the value of the criterion function obtained by moving di to one of the other k - 1 clusters. If there exist some moves that lead to an improvement in the overall value of the criterion function, then di is moved to the cluster that leads to the highest improvement." Each of the data points are evaluated during the training process. There is a threshold that is met and a score is generated. If there is a closer match to di then a new score is assigned and di will be moved to another cluster. This teaches that a new score is assigned based on a threshold. This process is executed iteratively, evaluating one to many documents, or transactions, in different clusters.)
Regarding claim 9, Zhao discloses, “identifying a congruent output label based on the comparing; and” Preliminaries, pp. 516; "The cosine formula can be simplified to
cos
d
i
,
d
j
=
d
i
t
d
j
, when the document vectors are of unit length. This measure becomes one if the documents are identical, and zero if there is nothing in common between them (i.e., the vectors are orthogonal to each other)." This system will evaluate the document for congruency. If the documents are the same or identical they will receive a specified value denoting so.)
“in response to the identifying the congruent output label, selecting an explainability score for the live transaction from the set of precomputed explanations.” (Criterion Function Optimization, pp. 518; "The refinement strategy that we used consists of a number of iterations. During each iteration, the documents are visited in a random order. For each document, di, we compute the change in the value of the criterion function obtained by moving di to one of the other k - 1 clusters.” During the training phases each of the clusters are visited. The data in the cluster is evaluated and a value is given to the data in the cluster. If and there is another cluster that better matches the data's score then that data is moved to another cluster. This data is document data which can be new data or previously evaluated data.)
Regarding claim 10, Zhao discloses, “wherein the selected explainability score is a unified explainability score for the most aligned cluster.” (Preliminaries, pp. 516; "The various clustering algorithms that are described in this paper use the vector-space model [24] to represent each document. In this model, each document d is considered to be a vector in the term-space." The model in this article is used to cluster documents. As stated above when clustering the model will determine a criterion based on the data type.) And (Criterion Function Optimization, pp. 518; “The greedy nature of the refinement algorithm does not guarantee that it will converge to a global minima, and the local minima solution it obtains depends on the particular set of seed documents that were selected during the initial clustering. To eliminate some of this sensitivity, the overall process is repeated a number of times. That is, we compute N different clustering solutions (i.e., initial clustering followed by cluster refinement), and the one that achieves the best value for the particular criterion function is kept.” As stated above, this model is able evaluate and score clusters which results in clusters of varying closeness. From this naturally a most aligned is produce and stored to be retrieved at a later time.)
Regarding claim 11, Zhao discloses, “updating the most aligned cluster to include the input features of the live transaction.” (Criterion Function Optimization, pp. 518; "If there exist some moves that lead to an improvement in the overall value of the criterion function, then di is moved to the cluster that leads to the highest improvement. If no such cluster exists, di remains in the cluster that it already belongs to. The refinement phase ends, as soon as we perform an iteration in which no documents moved between clusters." Each of the clusters are evaluated. While being evaluated, if there is another cluster with a better match, di is moved to another cluster. This teaches that the most aligned and the least aligned clusters are evaluated on certain criteria for closeness. The documents are embedded into the system which would include the documents features.)
Regarding claim 12, Dalli discloses, “wherein the set of labelled transactions comprises units of work performed within the database management system against a database.” (Detailed Description, pp. 38, [0435]; “In another exemplary embodiment, AutoXAI may be integrated with a workflow system with a bi-directional exchange of information between the AutoXAI system and the workflow system, including both processing data and event data. Symbolic information within the AutoXAI system may be made accessible to the workflow system, which can then take automated action accordingly. Conversely, workflow system information and workflow transition states or operational states may be made accessible to the AutoXAI system for control and configuration purposes. It is further contemplated that the AutoXAI and workflow combination may be further integrated within a Robotic Process Automation (RPA) system, Decision Support System (DSS) or a Data Lake system.” The system in this application is modular and able to work with many systems. This includes databased which contain data pertaining to calculations, queries, output data from various system, etc.)
Regarding claim 14, Zhao discloses, “wherein the selecting the homogeneous region comprises selecting a branch level having a homogeneity score above an adjustable threshold homogeneity score.” (Cluster Selection Schemes, pp. 518; "The key parameter in agglomerative algorithms is the method used to determine the pairs of clusters to be merged at each step. In most agglomerative algorithms, this is accomplished by selecting the most similar pair of clusters, and numerous approaches have been developed for computing the similarity between two clusters [25, 17, 14, 10, 11, 16]. In our study we used the single-link, complete-link, and UPGMA schemes, as well as, the various partitional criterion functions described in Section 3.1. The single-link [25] scheme measures the similarity of two clusters by the maximum similarity between the documents from each cluster. That is, the similarity between two clusters Si and Sj is given by [See Equation 8]." During the training process the data is evaluated and placed into clusters based on similarity. The similarity score can be used to understand the difference between clusters. The different clusters contain values which can be compared and placed accordingly. This system uses closeness criterion to determine the closeness of documents based on generated closeness values.)
Regarding claim 15, Dalli discloses, “A system for outputting explanations for live transactions implemented by a database management system, comprising: a memory; and a processor communicatively coupled to the memory, wherein the processor is configured to perform a method comprising:” (Detailed Description, pp. 35, [0420-0421]; “An exemplary embodiment may utilize different hardware for the implementation, including but not limited to: (i) Application Specific Integrated Circuits (ASICs), (ii) Field Programmable Gate Arrays (FPGAs), (iii) neuromorphic hardware or (iv) analogue/digital circuitry. [0421] Hardware may be used for a partial or full implementation of the AutoXAI system, either involving a complete self-contained system that may be used to perform AutoXAI on the device itself via its own dedicated interface, or by providing support for the user interface that may then be augmented by appropriate software and/or external information.” The system in this application is able to execute using different forms of hardware. This includes generic computing devices which contain a processor connected to memory which contains instructions to execute the method stored within.)
“generating, by the processor, an explainable artificial intelligence (XAI) module, wherein the generating comprises: obtaining a set of labeled transactions completed by the database management system, the labeled transactions comprising input features and corresponding output labels generated by a machine learning (ML) model;” (Detailed Description, pp. 3, [0045]; “An exemplary embodiment may implement an explainable neural network (which may be considered a form of XAI). Referring now to exemplary FIG. 12, FIG. 12 may illustrate a schematic diagram of an exemplary high level XNN architecture.” The model in this application is able to generate a XAI modules to be used in a multi-modal system. Fig. 12 is an example of a XNN module.) and (Detailed Description, pp. 4, [0062]; “Hierarchical clustering techniques or other logically equivalent methods can be used for identifying suitable partitions, such as an XAI model induction method which may input a set of training data to a black-box system and analyze the corresponding output using a partitioning function.” The XAI module in this application is able to evaluate input data and output data from another ML model. Further the system is able to cluster data using hierarchical techniques.)
“hierarchically clustering an input space of the labeled transactions based on the input features to form input space clusters of the labeled transactions divided into branch levels;” (Detailed Description, pp. 4, [0064]; “The input to the partitioning method can be either the input features directly for low-dimensional data (i.e., tabular) or data which has been pre-processed (for example, from a convolutional network). Features which have been transformed using a convolutional process may typically represent a higher-level of abstraction such as an edge, a stroke, or a pattern.” The system in this application is able to cluster, or partition, data using hierarchal clustering. This system will cluster data based on the input data and model output data.) and (Detailed Description, pp. 4, [0063]; “In an exemplary embodiment that may use a hierarchical clustering method for partitioning, a variety of appropriate methods may be used for a practical implementation including, but not limited to, agglomerative clustering, divisive clustering, relocation partitioning, probabilistic clustering, k-medoid methods, k-means methods, fuzzy clustering, density based clustering, grid based methods, gradient descent based methods, evolutionary methods, region splitting, region growing, sub-space clustering, projection methods, co-clustering methods and lazy clustering methods.” The proposed system is able to perform many different forms of hierarchical clustering. As disclosed, agglomerative clustering may be used which utilizes a tree structure containing branches and nodes/leaf.)
“obtaining, by an explainability model, explainability scores for labeled transactions in the homogeneous region;” (Detailed Description, pp. 14, [0189]; “A partition is a region in the data, which may be disjointing or overlapping. A rule may be a linear or nonlinear equation which may include coefficients with their respective dimension, and the result may represent both the answer to the problem and the explanation coefficients which may be used to generate domain specific explanations that are both machine and human readable. A rule may further represent a justification that explains how the explanation itself was produced. An exemplary embodiment applies an element of human readability to the encoded knowledge, data and rules which are otherwise too complex for an ordinary person to reproduce or comprehend without any automated process.” The model in this application is able to establish rules or a set of rules pertaining to the final explanation. This process can be performed on different partitions or clusters generated earlier. As stated above, a rule can be derived from a mathematical transformation or equation. Further, each of the partitions or spaces containing similar features can include assigned values which is used to generate explanations.)
“generating a unified explainability score for each input space cluster based on the explainability scores; and” (Detailed Description, pp. 14, [0186-0187]; “[0186] Still referring to exemplary FIG. 16, the method may continue by aggregating the data point predictions or classifications into hierarchal partitions 208. Rule conditions may be obtained from the hierarchal partitions. An external function defined by Partition(X) may identify the partitions. Partition(X) may be a function configured to partition similar data and may be used to create rules. The partitioning function may include a clustering algorithm such as k-means, entropy, or a mutual information-based method. [0187] The hierarchical partitions may organize the output data points in a variety of ways. In an exemplary embodiment, the data points may be aggregated such that each partition represents a rule or a set of rules. The hierarchical partitions may then be modeled using mathematical transformations and linear models.” As stated above, each partition or cluster can have rules applied. These rules can perform mathematical functions and are able to generate values for a partition. The partition or clusters are designed to group similar features to space or region using rules or a set of rules. This information will be used to generate an explanation.)
“storing each unified explainability score in a set of precomputed explanations; and” (Detailed Description, pp. 3, [0046]; “The layers may be analyzed by the selection and ranking layer 128 that may multiply the switch output by the value output, producing a ranked or scored output 130. The explanations and answers may be concurrently calculated by the XNN by the conditional network and the prediction network. The selection and ranking layer 128 may ensure that the answers and explanations are correctly matched, ranked and scored appropriately before being sent to the output 130.” The model in this application is part of a multi-model system. As stated above, some components or models run concurrently. The actions of some models may be completed and stored which are retrieved and evaluated at a later point to output an explanation.)
“selecting, by the processor using the XAI module, from the set of precomputed explanations, an explainability score for a live transaction.” (Detailed Description, pp. 7, [0099]; “The end result 914 of the explanation process may include either an explanation and/or its interpretation, which may be consumed by a human user, another application, another system component forming part of a larger embodiment, or some other automated system.” The model in this application is able to output an explanation based on input information.) and (Detailed Description, pp. 8, [0104-0105]; “The explanation scaffolding 9101 can also store audit data in appropriate components such as the actions 1538 or the evaluation 1535 components, or via an independent audit system plugged in to the third-party data 1539 extension component. [0105] The interpretation scaffolding 9111, illustrated in FIG. 15C, can also store audit data in appropriate components such as the interactive context 1544 or the protocol context 1548 components. The interpretation scaffolding 9111 may be structured into three main components, the explanation and interpretation scenario component 9112, the framing, protocol, and contextual component 1540, and the interpretation model component 1550.” This model also uses stored values from previous explanations to produce current explanations. This teaches that the output of this model is generated based on or uses prior computer values and outcomes.)
Dalli fails to explicitly disclose, “identifying a homogenous region of the input space, wherein the identifying comprises: scoring homogeneity of the input space clusters based on a number of matching output labels from the output labels corresponding to the labeled transactions in each input space cluster;”, “scoring homogeneity of the branch levels based on the homogeneity scoring of the input space clusters; and” and “selecting the homogeneous region based on the homogeneity scoring of the branch levels;”.
However, Zhao discloses, “identifying a homogenous region of the input space, wherein the identifying comprises: scoring homogeneity of the input space clusters based on a number of matching output labels from the output labels corresponding to the labeled transactions in each input space cluster;” (Internal Criterion Functions, pp. 517; “The first internal criterion function maximizes the sum of the average pairwise similarities between the documents assigned to each cluster, weighted according to the size of each cluster. Specifically, if we use the cosine function to measure the similarity between documents, then we want the clustering solution to minimize the following criterion function: [See Equation (3)].” Zhao discloses different methods of clustering and gives a description to the different scoring criterion used in clustering documents using hierarchical clustering. This teaches the evaluation of cluster to determine similarity in the cluster. This is used to generate a homogenous cluster.)
“scoring homogeneity of the branch levels based on the homogeneity scoring of the input space clusters; and” (Internal Criterion Functions, pp. 517; “The second criterion function is used by the popular vector-space variant of the K-means algorithm [6, 18, 7, 26, 15]. In this algorithm each cluster is represented by its centroid vector and the goal is to find the clustering solution that maximizes the similarity between each document and the centroid of the cluster that is assigned to. Specifically, if we use the cosine function to measure the similarity between a document and a centroid, then the criterion function maximizes the following: [See Equation (4)].” Zhao discloses further criterion for clustering which involve evaluating a clusters content to the center of the cluster. This teaches an evaluation of the different branch levels or clusters.)
“selecting the homogeneous region based on the homogeneity scoring of the branch levels;” (External Criterion Functions, pp. 517; “As we can see from Equation 5, even-though our initial motivation was to define an external criterion function, because we used the cosine function to measure the separation between the cluster and the entire collection, the criterion function does take into account the within-cluster similarity of the documents (due to the ||Dr|| term). Thus, ℇ1 is actually a hybrid criterion function that combines both external and internal characteristics of the clusters.” Zhao discloses a process of evaluating a cluster or branch to the whole model. This is disclosed as an external criterion. This process including selecting a cluster and evaluating it.)
Regarding claim 16, Dalli discloses, “receiving the live transaction from a user device; and” (Detailed Description, pp. 37, [0434]; “In another exemplary embodiment, an AutoXAI may be integrated with a Robotic Process Automation (RPA) system with a bi-directional exchange of information between the AutoXAI system and the RPA system. Symbolic information within the AutoXAI system may be made accessible to the RPA system, which can then take automated action accordingly.” The process disclosed in Dalli can be applied to various different models. This can include, as stated above, a robotic system. This system is able to work online and live able to take and evaluate transactions provided in updates or communication.)
“extracting, by the XAI module, input features from the live transaction.” (Detailed Description, pp. 4, [0064]; “The input to the partitioning method can be either the input features directly for low-dimensional data (i.e., tabular) or data which has been pre-processed (for example, from a convolutional network). Features which have been transformed using a convolutional process may typically represent a higher-level of abstraction such as an edge, a stroke, or a pattern.” The model in this application is able to extract features from transactions or previously computed output data from a machine learning model.)
Regarding claim 17, Dalli discloses, “wherein the explainability score is selected based on the input features of the live transaction.” (Detailed Description, pp. 36, [0427]; “In parts of the flight envelope, some input may lead to dangerous instability or even irrecoverable situations, such as "deep stall" or "flat spin", respectively. An AI system may be able to analyze these situations in order to prevent them from happening. However, an exemplary XAI system may be able to explain which input configurations are causing dangerous flight conditions and offer suggested changes as part of the explanation, so that the pilot may quickly adjust the control stick input to recover good flight characteristics or, if deemed necessary, retain the current input.” The XAI system in this application is designed to be modular. As stated above this system can run in real time and produce explanations based on regular inputs or transactions from an active system.)
Regarding claim 18, Dalli discloses, “A computer program product, the computer program product comprising a computer readable storage medium having program instructions embodied therewith, the program instructions executable by a processor to cause a device to perform a method of outputting explanations for live transactions implemented by a database management system, the method comprising:” (Detailed Description, pp. 35, [0420-0421]; “An exemplary embodiment may utilize different hardware for the implementation, including but not limited to: (i) Application Specific Integrated Circuits (ASICs), (ii) Field Programmable Gate Arrays (FPGAs), (iii) neuromorphic hardware or (iv) analogue/digital circuitry. [0421] Hardware may be used for a partial or full implementation of the AutoXAI system, either involving a complete self-contained system that may be used to perform AutoXAI on the device itself via its own dedicated interface, or by providing support for the user interface that may then be augmented by appropriate software and/or external information.” The system in this application is able to execute using different forms of hardware. This includes generic computing devices which contain a processor connected to memory which contains instructions to execute the method stored within.)
“generating, by the processor, an explainable artificial intelligence (XAI) module, wherein the generating comprises: obtaining, a set of labeled transactions completed by the database management system, the labeled transactions comprising input features and corresponding output labels generated by a machine learning (ML) model;” (Detailed Description, pp. 3, [0045]; “An exemplary embodiment may implement an explainable neural network (which may be considered a form of XAI). Referring now to exemplary FIG. 12, FIG. 12 may illustrate a schematic diagram of an exemplary high level XNN architecture.” The model in this application is able to generate a XAI modules to be used in a multi-modal system. Fig. 12 is an example of a XNN module.) and (Detailed Description, pp. 4, [0062]; “Hierarchical clustering techniques or other logically equivalent methods can be used for identifying suitable partitions, such as an XAI model induction method which may input a set of training data to a black-box system and analyze the corresponding output using a partitioning function.” The XAI module in this application is able to evaluate input data and output data from another ML model. Further the system is able to cluster data using hierarchical techniques.)
“hierarchically clustering an input space of the labeled transactions based on the input features to form input space clusters of the labeled transactions divided into branch levels;” (Detailed Description, pp. 4, [0064]; “The input to the partitioning method can be either the input features directly for low-dimensional data (i.e., tabular) or data which has been pre-processed (for example, from a convolutional network). Features which have been transformed using a convolutional process may typically represent a higher-level of abstraction such as an edge, a stroke, or a pattern.” The system in this application is able to cluster, or partition, data using hierarchal clustering. This system will cluster data based on the input data and model output data.) and (Detailed Description, pp. 4, [0063]; “In an exemplary embodiment that may use a hierarchical clustering method for partitioning, a variety of appropriate methods may be used for a practical implementation including, but not limited to, agglomerative clustering, divisive clustering, relocation partitioning, probabilistic clustering, k-medoid methods, k-means methods, fuzzy clustering, density based clustering, grid based methods, gradient descent based methods, evolutionary methods, region splitting, region growing, sub-space clustering, projection methods, co-clustering methods and lazy clustering methods.” The proposed system is able to perform many different forms of hierarchical clustering. As disclosed, agglomerative clustering may be used which utilizes a tree structure containing branches and nodes/leaf.)
“obtaining, by an explainability model, explainability scores for labeled transactions in the homogeneous region;” (Detailed Description, pp. 14, [0189]; “A partition is a region in the data, which may be disjointing or overlapping. A rule may be a linear or nonlinear equation which may include coefficients with their respective dimension, and the result may represent both the answer to the problem and the explanation coefficients which may be used to generate domain specific explanations that are both machine and human readable. A rule may further represent a justification that explains how the explanation itself was produced. An exemplary embodiment applies an element of human readability to the encoded knowledge, data and rules which are otherwise too complex for an ordinary person to reproduce or comprehend without any automated process.” The model in this application is able to establish rules or a set of rules pertaining to the final explanation. This process can be performed on different partitions or clusters generated earlier. As stated above, a rule can be derived from a mathematical transformation or equation. Further, each of the partitions or spaces containing similar features can include assigned values which is used to generate explanations.)
“generating a unified explainability score for each input space cluster in the homogeneous region based on the explainability scores; and” (Detailed Description, pp. 14, [0186-0187]; “[0186] Still referring to exemplary FIG. 16, the method may continue by aggregating the data point predictions or classifications into hierarchal partitions 208. Rule conditions may be obtained from the hierarchal partitions. An external function defined by Partition(X) may identify the partitions. Partition(X) may be a function configured to partition similar data and may be used to create rules. The partitioning function may include a clustering algorithm such as k-means, entropy, or a mutual information-based method. [0187] The hierarchical partitions may organize the output data points in a variety of ways. In an exemplary embodiment, the data points may be aggregated such that each partition represents a rule or a set of rules. The hierarchical partitions may then be modeled using mathematical transformations and linear models.” As stated above, each partition or cluster can have rules applied. These rules can perform mathematical functions and are able to generate values for a partition. The partition or clusters are designed to group similar features to space or region using rules or a set of rules. This information will be used to generate an explanation.)
“storing each unified explainability score in a set of precomputed explanations; and” (Detailed Description, pp. 3, [0046]; “The layers may be analyzed by the selection and ranking layer 128 that may multiply the switch output by the value output, producing a ranked or scored output 130. The explanations and answers may be concurrently calculated by the XNN by the conditional network and the prediction network. The selection and ranking layer 128 may ensure that the answers and explanations are correctly matched, ranked and scored appropriately before being sent to the output 130.” The model in this application is part of a multi-model system. As stated above, some components or models run concurrently. The actions of some models may be completed and stored which are retrieved and evaluated at a later point to output an explanation.)
“selecting, by the processor using the XAI module, from the set of precomputed explanations, an explainability score for a live transaction.” (Detailed Description, pp. 7, [0099]; “The end result 914 of the explanation process may include either an explanation and/or its interpretation, which may be consumed by a human user, another application, another system component forming part of a larger embodiment, or some other automated system.” The model in this application is able to output an explanation based on input information.) and (Detailed Description, pp. 8, [0104-0105]; “The explanation scaffolding 9101 can also store audit data in appropriate components such as the actions 1538 or the evaluation 1535 components, or via an independent audit system plugged in to the third-party data 1539 extension component. [0105] The interpretation scaffolding 9111, illustrated in FIG. 15C, can also store audit data in appropriate components such as the interactive context 1544 or the protocol context 1548 components. The interpretation scaffolding 9111 may be structured into three main components, the explanation and interpretation scenario component 9112, the framing, protocol, and contextual component 1540, and the interpretation model component 1550.” This model also uses stored values from previous explanations to produce current explanations. This teaches that the output of this model is generated based on or uses prior computer values and outcomes.)
Dalli fails to explicitly disclose, “identifying a homogenous region of the input space, wherein the identifying comprises: scoring homogeneity of the input space clusters based on a number of matching output labels from the corresponding output labels corresponding to the labeled transactions in each input space cluster;”, “scoring homogeneity of the branch levels based on the homogeneity scoring of the input space clusters; and” and “selecting the homogeneous region based on the homogeneity scoring of the branch levels;”.
However, Zhao discloses, “identifying a homogenous region of the input space, wherein the identifying comprises: scoring homogeneity of the input space clusters based on a number of matching output labels from the corresponding output labels corresponding to the labeled transactions in each input space cluster;” (Internal Criterion Functions, pp. 517; “The first internal criterion function maximizes the sum of the average pairwise similarities between the documents assigned to each cluster, weighted according to the size of each cluster. Specifically, if we use the cosine function to measure the similarity between documents, then we want the clustering solution to minimize the following criterion function: [See Equation (3)].” Zhao discloses different methods of clustering and gives a description to the different scoring criterion used in clustering documents using hierarchical clustering. This teaches the evaluation of cluster to determine similarity in the cluster. This is used to generate a homogenous cluster.)
“scoring homogeneity of the branch levels based on the homogeneity scoring of the input space clusters; and” (Internal Criterion Functions, pp. 517; “The second criterion function is used by the popular vector-space variant of the K-means algorithm [6, 18, 7, 26, 15]. In this algorithm each cluster is represented by its centroid vector and the goal is to find the clustering solution that maximizes the similarity between each document and the centroid of the cluster that is assigned to. Specifically, if we use the cosine function to measure the similarity between a document and a centroid, then the criterion function maximizes the following: [See Equation (4)].” Zhao discloses further criterion for clustering which involve evaluating a clusters content to the center of the cluster. This teaches an evaluation of the different branch levels or clusters.)
“selecting the homogeneous region based on the homogeneity scoring of the branch levels;” (External Criterion Functions, pp. 517; “As we can see from Equation 5, even-though our initial motivation was to define an external criterion function, because we used the cosine function to measure the separation between the cluster and the entire collection, the criterion function does take into account the within-cluster similarity of the documents (due to the ||Dr|| term). Thus, ℇ1 is actually a hybrid criterion function that combines both external and internal characteristics of the clusters.” Zhao discloses a process of evaluating a cluster or branch to the whole model. This is disclosed as an external criterion. This process including selecting a cluster and evaluating it.)
Regarding claim 19, Dalli discloses, “receiving the live transaction from a user device; and” (Detailed Description, pp. 37, [0434]; “In another exemplary embodiment, an AutoXAI may be integrated with a Robotic Process Automation (RPA) system with a bi-directional exchange of information between the AutoXAI system and the RPA system. Symbolic information within the AutoXAI system may be made accessible to the RPA system, which can then take automated action accordingly.” The process disclosed in Dalli can be applied to various different models. This can include, as stated above, a robotic system. This system is able to work online and live able to take and evaluate transactions provided in updates or communication.)
“extracting, by the XAI module, input features from the live transaction.” (Detailed Description, pp. 4, [0064]; “The input to the partitioning method can be either the input features directly for low-dimensional data (i.e., tabular) or data which has been pre-processed (for example, from a convolutional network). Features which have been transformed using a convolutional process may typically represent a higher-level of abstraction such as an edge, a stroke, or a pattern.” The model in this application is able to extract features from transactions or previously computed output data from a machine learning model.)
Regarding claim 20, Dalli discloses, “wherein the explainability score is selected based on the input features of the live transaction.” (Detailed Description, pp. 36, [0427]; “In parts of the flight envelope, some input may lead to dangerous instability or even irrecoverable situations, such as "deep stall" or "flat spin", respectively. An AI system may be able to analyze these situations in order to prevent them from happening. However, an exemplary XAI system may be able to explain which input configurations are causing dangerous flight conditions and offer suggested changes as part of the explanation, so that the pilot may quickly adjust the control stick input to recover good flight characteristics or, if deemed necessary, retain the current input.” The XAI system in this application is designed to be modular. As stated above this system can run in real time and produce explanations based on regular inputs or transactions from an active system.)
Regarding claim 21, Zhao discloses, “identifying a most aligned cluster from the homogeneous region based on input features of the live transaction; and” (Clustering, pp. 8; "The refinement strategy that we used consists of a number of iterations. During each iteration, the documents are visited in a random order. For each document, di, we compute the change in the value of the criterion function obtained by moving di to one of the other k - 1 clusters. If there exist some moves that lead to an improvement in the overall value of the criterion function, then di is moved to the cluster that leads to the highest improvement." During training each of the pairs are selected. This can include the most aligned and least aligned cluster to look for improvements. This process is executed iteratively, evaluating one to many documents, or transactions, in different clusters.)
“determining that the most aligned cluster satisfies a closeness criterion;” (Criterion Function Optimization, pp. 518; "If there exist some moves that lead to an improvement in the overall value of the criterion function, then di is moved to the cluster that leads to the highest improvement. If no such cluster exists, di remains in the cluster that it already belongs to. The refinement phase ends, as soon as we perform an iteration in which no documents moved between clusters." Each of the clusters are evaluated. While being evaluated if there is another cluster with a better match di is moved to another cluster. This teaches that the most aligned and the least aligned clusters are evaluated on certain criteria for closeness.)
“in response to determining that the most aligned cluster satisfies the closeness criterion, comparing output labels of the most aligned cluster with an output label of the live transaction.” (Preliminaries, pp. 516; "The various clustering algorithms that are described in this paper use the vector-space model [24] to represent each document. In this model, each document d is considered to be a vector in the term-space." The model in this article is used to cluster documents. As stated above when clustering the model will determine a criterion based on the data type.) and (Criterion Function Optimization, pp. 518; “Initially, a random pair of documents is selected from the collection to act as the seeds of the two clusters. Then, for each document, its similarity to these two seeds is computed, and it is assigned to the cluster corresponding to its most similar seed. This forms the initial two-way clustering. This clustering is then repeatedly refined so that it optimizes the desired clustering criterion function.” This system will start with two documents and compare them to each other. If the documents are similar, the values are mapped to each other in the cluster. As each document is added the documents of the nodes are compared and evaluated for closeness.)
Regarding claim 22, Dalli discloses, “wherein the generating the XAI module further comprises storing the explainability scores for the labeled transactions in the set of precomputed explanations.” (Detailed Description, pp. 8, [0104-0106]; “The explanation scaffolding 9101 can also store audit data in appropriate components such as the actions 1538 or the evaluation 1535 components, or via an independent audit system plugged in to the third-party data 1539 extension component. [0105] The interpretation scaffolding 9111, illustrated in FIG. 15C, can also store audit data in appropriate components such as the interactive context 1544 or the protocol context 1548 components. The interpretation scaffolding 9111 may be structured into three main components, the explanation and interpretation scenario component 9112, the framing, protocol, and contextual component 1540, and the interpretation model component 1550. [0106] The explanation and interpretation scenario component 9112 may include a suitable version of the explanation scaffolding 9101, such as the one illustrated in FIG. 15B, together with an interpretation scenario 9113. The interpretation scenario 9113 may be used to aid in the creation and configuration of an interpretation brief 1547. The interpretation scenario 9113 may optionally influence the behavior of the selection model 1553.” The system in this application is able to produce an explanation based on the input values. When producing the final explanation, the system is able to generate the explanation using prior data or explanations. The explanations can be produced using premade templates or scaffolding.)
Conclusion
THIS ACTION IS MADE FINAL. Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any nonprovisional extension fee (37 CFR 1.17(a)) pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to PAUL MICHAEL GALVIN-SIEBENALER whose telephone number is (571)272-1257. The examiner can normally be reached Monday - Friday 8AM to 5PM.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Viker Lamardo can be reached at (571) 270-5871. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of published or unpublished applications may be obtained from Patent Center. Unpublished application information in Patent Center is available to registered users. To file and manage patent submissions in Patent Center, visit: https://patentcenter.uspto.gov. Visit https://www.uspto.gov/patents/apply/patent-center for more information about Patent Center and https://www.uspto.gov/patents/docx for information about filing in DOCX format. For additional questions, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/PAUL M GALVIN-SIEBENALER/Examiner, Art Unit 2147
/ERIC NILSSON/Primary Examiner, Art Unit 2151