DETAILED ACTION
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Examiner Remarks
Documents or prior art in this document are identified as:
Weast et al (US 20210117864), referred to as Weast in this document,
Neal et al (US20220358240), referred to as Neal in this document,
Nowruzi et al ("How much real data do we actually need: Analyzing object detection performance using synthetic and real data"), referred to as Nowruzi in this document,
Redlich et al (US10250640), referred to as Redlich in this document.
Claim Objections
Claim 3 recites “includes at least one of (i) a global application; and (ii) a specific application.” which appears to contain a typo splitting the list. Claim 3 is interpreted with the semicolon (between “application” and “and”) removed.
Claim 34 recites “causes the computing device to carry out a method build a customized training dataset for training an artificial intelligence (Al) platform”. The recitation appears to contain a typo of “a method to build”.
Claim Rejections - 35 USC § 112
Regarding 112(b):
The following is a quotation of 35 U.S.C. 112(b):
(b) CONCLUSION.—The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the inventor or a joint inventor regards as the invention.
The following is a quotation of 35 U.S.C. 112 (pre-AIA ), second paragraph:
The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the applicant regards as his invention.
Claims 12, 14, 15, 17, 20-24 rejected under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), second paragraph, as being indefinite for failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor (or for applications subject to pre-AIA 35 U.S.C. 112, the applicant), regards as the invention.
In regard to Claim 12:
Claim 12 recites the limitation "the training datasets" in “wherein the training datasets each have a common format”. There is insufficient antecedent basis for this limitation in the claim. One of ordinary skill would be unsure which training datasets are being referred to (as claim 1 includes both training datasets in a database and candidate training datasets). If claim 12 is referring to training datasets as a format, then a possible alteration could say “wherein training datasets” to refer to all training datasets rather than referring to a specific collection of training datasets.
In regard to claim 17:
Claim 17 recites the limitation "the media metadata in the third data block". There is insufficient antecedent basis for this limitation in the claim. “Media metadata” and “third data block” are elements that have not been introduced. Claim 12, which claim 17 does not depend on, is the claim that introduces such elements.
Claim 17 recites the limitation “the training datasets” in " comparing the specifications to the media metadata in the third data block of each of the training datasets" and “including those training datasets for which the media metadata in the third data block of the respective training datasets matches the specifications”. There is insufficient antecedent basis for this limitation in the claim. One of ordinary skill could not determine which training datasets are being referred to within the limitation. Claim 1 introduces a database of training datasets and a candidate set of training datasets.
In regard to claim 20:
Claim 20 recites the limitation "the second data block". There is insufficient antecedent basis for this limitation in the claim. “Second data block” containing use rights are elements that have not been introduced. Claim 12, which claim 20 does not depend on, is the claim that introduces such elements.
Claim 20 recites the limitation “the respective training dataset”. There is insufficient antecedent basis for this limitation in the claim. One of ordinary skill could not determine which training datasets are being referred to within the limitation. Claim 1 introduces a database of training datasets and a candidate set of training datasets.
Claim 20 recites the limitation "a candidate set of training datasets" in” wherein to identify a candidate set of training datasets matching the specifications”. There is insufficient antecedent basis for this limitation in the claim. “a candidate set of training datasets” was already introduced in claim 1, thus one cannot tell if this limitation introduces a new dataset or refers to the existing candidate set.
In regard to claim 21:
Claim 21 recites the limitation “the first data block” and "the third data block". There is insufficient antecedent basis for this limitation in the claim. “third data block” containing a list of objects and “the media metadata corresponding to the training media in the first data block” are elements that have not been introduced. Claim 12, which claim 21 does not depend on, is the claim that introduces such elements.
Claim 20 recites the limitation “the training dataset” and “the respective training dataset” in “the training datasets, the candidate set of training datasets excluding those training datasets for which the list of objects in the third data block of the respective training dataset”. There is insufficient antecedent basis for this limitation in the claim. One of ordinary skill could not determine which training datasets are being referred to within the limitation. Claim 1 introduces a database of training datasets and a candidate set of training datasets.
Claim 21 recites the limitation "a candidate set of training datasets" in ”wherein to identify a candidate set of training datasets matching the specifications”. There is insufficient antecedent basis for this limitation in the claim. “a candidate set of training datasets” was already introduced in claim 1, thus one cannot tell if this limitation introduces a new dataset or refers to the existing candidate set.
In regard to dependent claims:
Claims dependent upon claims rejected under 112(b) are rejected for being dependent upon a rejected claim.
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 1-3, 5, 8-12, 14, 17, 20-24, 34 are rejected under 35 U.S.C. 101 because the claimed invention is directed towards an abstract idea without significantly more.
In regards to Claim 1:
Step 1: Is the claim directed towards a process, machine, manufacture, or composition of matter?
Yes, the claim is directed towards a method, so a process.
Step 2A Prong 1: Does the claim recite a law of nature, a natural phenomenon, or an abstract idea?
Yes, the claim does recite a(n) abstract idea.
Claim 1 recites the following abstract ideas:
to identify a candidate set of training datasets matching the specifications
This limitation is directed towards the abstract idea of a mental process, or a concept performed in the human mind, including observation, evaluation, judgement or opinion (see MPEP 2106.04(a)(2) subsection 3). Here the limitation is seen as evaluation and judgement.
and authorizing release of a subset of the training datasets in the candidate set of training datasets based on the data indicative of the proposed usage of the Al training data and the use rights associated with the training data sets in the candidate set of training datasets
This limitation is directed towards the abstract idea of a mental process, or a concept performed in the human mind, including observation, evaluation, judgement or opinion (see MPEP 2106.04(a)(2) subsection 3). Here the limitation is seen as judgement.
Step 2A Prong 2: Does the claim recite additional elements that integrate the exception into a practical application of the exception?
No, the application does not recite any additional elements that would integrate the abstract idea into a practical application.
Claim 1 recites the following additional elements:
A method of operating a computer to build a customized training dataset for training an artificial intelligence (Al) platform, the method comprising
At a high level of generality, this is an activity of using a computer as an “apply it” use (see MPEP 2106.05(f)).
obtaining (i) specifications of training data for training the Al platform and (ii) data indicative of a proposed usage of the training data
This limitation is directed towards the insignificant extra solution activity of mere data gathering (see MPEP § 2106.05(g)).
consulting a database of training datasets, each associated with use rights
This limitation is directed towards the insignificant extra solution activity of mere data gathering (see MPEP § 2106.05(g))
and authorizing release of a subset of the training datasets in the candidate set of training datasets based on the data indicative of the proposed usage of the Al training data and the use rights associated with the training data sets in the candidate set of training datasets
This limitation is directed towards the insignificant extra solution activity of mere data gathering (see MPEP § 2106.05(g)).
Step 2B: Does the claim as a whole amount to significantly more than the judicial exception?
No, the claim as a whole does not amount to significantly more than the judicial exception. All elements of the claim, viewed individually or wholistically, do not provide an inventive concept or otherwise significantly more than the abstract idea itself.
Claim 1 recites the following additional elements:
A method of operating a computer to build a customized training dataset for training an artificial intelligence (Al) platform, the method comprising
At a high level of generality, this is an activity of using a computer as an “apply it” use (see MPEP 2106.05(f)). At said high level of generality, a computer or computer parts appears to be an implementation of the abstract idea on a computer, so merely using a computer as a tool to perform the abstract idea.
obtaining (i) specifications of training data for training the Al platform and (ii) data indicative of a proposed usage of the training data
This limitation is directed towards the insignificant extra solution activity of mere data gathering (see MPEP § 2106.05(g)). This is a well understood, routine, conventional activity of transmitting data (see MPEP 2106.05(d) example i in computer functions).
consulting a database of training datasets, each associated with use rights
This limitation is directed towards the insignificant extra solution activity of mere data gathering (see MPEP § 2106.05(g)). This is a well understood, routine, conventional activity of transmitting data (see MPEP 2106.05(d) example i in computer functions).
and authorizing release of a subset of the training datasets in the candidate set of training datasets based on the data indicative of the proposed usage of the Al training data and the use rights associated with the training data sets in the candidate set of training datasets
This limitation is directed towards the insignificant extra solution activity of mere data gathering (see MPEP § 2106.05(g)). This is a well understood, routine, conventional activity of transmitting data (see MPEP 2106.05(d) example i in computer functions).
In regards to Claim 2:
Step 2A Prong 2: Does the claim recite additional elements that integrate the exception into a practical application of the exception?
No, the application does not recite any additional elements that would integrate the abstract idea into a practical application.
Claim 2 recites the following additional elements:
wherein the specifications include an application scope
This limitation is directed towards a continuation of the insignificant extra solution activity of mere data gathering (see MPEP § 2106.05(g)).
Step 2B: Does the claim as a whole amount to significantly more than the judicial exception?
No, the claim as a whole does not amount to significantly more than the judicial exception. All elements of the claim, viewed individually or wholistically, do not provide an inventive concept or otherwise significantly more than the abstract idea itself.
Claim 2 recites the following additional elements:
wherein the specifications include an application scope
This limitation is directed towards a continuation of the insignificant extra solution activity of mere data gathering (see MPEP § 2106.05(g)). This is a well understood, routine, conventional activity of transmitting data (see MPEP 2106.05(d) example i in computer functions). Noting data that is included in the input does not integrate into a practical application.
In regards to Claim 3:
Step 2A Prong 2: Does the claim recite additional elements that integrate the exception into a practical application of the exception?
No, the application does not recite any additional elements that would integrate the abstract idea into a practical application.
Claim3 recites the following additional elements:
wherein the application scope includes at least one of (i) a global application; and (ii) a specific application
This limitation is directed towards a continuation of the insignificant extra solution activity of mere data gathering (see MPEP § 2106.05(g)).
Step 2B: Does the claim as a whole amount to significantly more than the judicial exception?
No, the claim as a whole does not amount to significantly more than the judicial exception. All elements of the claim, viewed individually or wholistically, do not provide an inventive concept or otherwise significantly more than the abstract idea itself.
Claim 3 recites the following additional elements:
wherein the application scope includes at least one of (i) a global application; and (ii) a specific application
This limitation is directed towards a continuation of the insignificant extra solution activity of mere data gathering (see MPEP § 2106.05(g)). This is a well understood, routine, conventional activity of transmitting data (see MPEP 2106.05(d) example i in computer functions). Noting data that is included in the input does not integrate into a practical application.
In regards to Claim 5:
Step 2A Prong 2: Does the claim recite additional elements that integrate the exception into a practical application of the exception?
No, the application does not recite any additional elements that would integrate the abstract idea into a practical application.
Claim 5 recites the following additional elements:
wherein the specifications include a minimum percentage of training data derived from real-world imaging or a maximum percentage of training data derived synthetically
This limitation is directed towards a continuation of the insignificant extra solution activity of mere data gathering (see MPEP § 2106.05(g)).
Step 2B: Does the claim as a whole amount to significantly more than the judicial exception?
No, the claim as a whole does not amount to significantly more than the judicial exception. All elements of the claim, viewed individually or wholistically, do not provide an inventive concept or otherwise significantly more than the abstract idea itself.
Claim 5 recites the following additional elements:
wherein the specifications include a minimum percentage of training data derived from real-world imaging or a maximum percentage of training data derived synthetically
This limitation is directed towards a continuation of the insignificant extra solution activity of mere data gathering (see MPEP § 2106.05(g)). This is a well understood, routine, conventional activity of transmitting data (see MPEP 2106.05(d) example i in computer functions). Noting data that is included in the input does not integrate into a practical application.
In regards to Claim 6:
Step 2A Prong 2: Does the claim recite additional elements that integrate the exception into a practical application of the exception?
No, the application does not recite any additional elements that would integrate the abstract idea into a practical application.
Claim 6 recites the following additional elements:
wherein the data indicative of a proposed usage of the training data includes one or more of (i) an identifier of the AI platform and (ii) a proposed geographic region
This limitation is directed towards a continuation of the insignificant extra solution activity of mere data gathering (see MPEP § 2106.05(g)).
Step 2B: Does the claim as a whole amount to significantly more than the judicial exception?
No, the claim as a whole does not amount to significantly more than the judicial exception. All elements of the claim, viewed individually or wholistically, do not provide an inventive concept or otherwise significantly more than the abstract idea itself.
Claim 6 recites the following additional elements:
wherein the data indicative of a proposed usage of the training data includes one or more of (i) an identifier of the AI platform and (ii) a proposed geographic region
This limitation is directed towards a continuation of the insignificant extra solution activity of mere data gathering (see MPEP § 2106.05(g)). This is a well understood, routine, conventional activity of transmitting data (see MPEP 2106.05(d) example i in computer functions). Noting data that is included in the input does not integrate into a practical application.
In regards to Claim 8:
Step 2A Prong 1: Does the claim recite a law of nature, a natural phenomenon, or an abstract idea?
Yes, the claim does recite a(n) abstract idea.
Claim 8 recites the following abstract ideas:
wherein the authorizing comprises creating a subset of the candidate set of training datasets, the subset including those training datasets in the candidate set for which the use rights include the proposed usage of the AI training data
This limitation is directed towards the continuation of the abstract idea in claim 1 of a mental process, or a concept performed in the human mind, including observation, evaluation, judgement or opinion (see MPEP 2106.04(a)(2) subsection 3). Here the limitation is seen as continuing the authorizing to note filtering or creating a subject utilizing judgement.
In regards to Claim 9:
Step 2A Prong 1: Does the claim recite a law of nature, a natural phenomenon, or an abstract idea?
Yes, the claim does recite a(n) abstract idea.
Claim 9 recites the following abstract ideas:
wherein the subset of the candidate set of training datasets excludes any training datasets in the candidate set for which the use rights do not include the proposed usage of the AI training data
This limitation is directed towards the continuation of the abstract idea in claim 1 and 8 of a mental process, or a concept performed in the human mind, including observation, evaluation, judgement or opinion (see MPEP 2106.04(a)(2) subsection 3). Here the limitation is seen as continuing the authorizing to note filtering or creating a subject utilizing judgement.
In regards to Claim 10:
Step 2A Prong 1: Does the claim recite a law of nature, a natural phenomenon, or an abstract idea?
Yes, the claim does recite a(n) abstract idea.
Claim 10 recites the following abstract ideas:
the method further comprising determining whether the user has a license to train the AI platform
This limitation is directed towards the abstract idea of a mental process, or a concept performed in the human mind, including observation, evaluation, judgement or opinion (see MPEP 2106.04(a)(2) subsection 3). Here the limitation is seen as evaluation.
Step 2A Prong 2: Does the claim recite additional elements that integrate the exception into a practical application of the exception?
No, the application does not recite any additional elements that would integrate the abstract idea into a practical application.
Claim 10 recites the following additional elements:
wherein the specifications and the data indicative of a proposed usage of the training data are obtained from a user
This limitation is directed towards the insignificant extra solution activity of mere data gathering (see MPEP § 2106.05(g)).
Step 2B: Does the claim as a whole amount to significantly more than the judicial exception?
No, the claim as a whole does not amount to significantly more than the judicial exception. All elements of the claim, viewed individually or wholistically, do not provide an inventive concept or otherwise significantly more than the abstract idea itself.
Claim 10 recites the following additional elements:
wherein the specifications and the data indicative of a proposed usage of the training data are obtained from a user
This limitation is directed towards the insignificant extra solution activity of mere data gathering (see MPEP § 2106.05(g)). This is a well understood, routine, conventional activity of transmitting data (see MPEP 2106.05(d) example i in computer functions).
In regards to Claim 11:
Step 2A Prong 1: Does the claim recite a law of nature, a natural phenomenon, or an abstract idea?
Yes, the claim does recite a(n) abstract idea.
Claim 11 recites the following abstract ideas:
and wherein if the user is determined to have a license to train the AI platform, the consulting comprises a step of decrypting the at least some of the training datasets in the database
This limitation is directed towards the abstract idea of a mental process, or a concept performed in the human mind, including observation, evaluation, judgement or opinion (see MPEP 2106.04(a)(2) subsection 3). Here the limitation is seen as evaluation and judgement. The broad premise of decrypting something given a criteria is performable in the human mind, such as a person deciding to tell someone what something means should the person have an appropriate ID.
Step 2A Prong 2: Does the claim recite additional elements that integrate the exception into a practical application of the exception?
No, the application does not recite any additional elements that would integrate the abstract idea into a practical application.
Claim 11 recites the following additional elements:
wherein at least some of the training datasets in the database are encrypted
This limitation is directed towards a continuation of the insignificant extra solution activity of mere data gathering (see MPEP § 2106.05(g)).
Step 2B: Does the claim as a whole amount to significantly more than the judicial exception?
No, the claim as a whole does not amount to significantly more than the judicial exception. All elements of the claim, viewed individually or wholistically, do not provide an inventive concept or otherwise significantly more than the abstract idea itself.
Claim 11 recites the following additional elements:
wherein at least some of the training datasets in the database are encrypted
This limitation is directed towards a continuation of the insignificant extra solution activity of mere data gathering (see MPEP § 2106.05(g)). This is a well understood, routine, conventional activity of transmitting data (see MPEP 2106.05(d) example i in computer functions). Noting data that is included in the input being encrypting in form does not integrate into a practical application.
In regards to Claim 12:
Step 2A Prong 2: Does the claim recite additional elements that integrate the exception into a practical application of the exception?
No, the application does not recite any additional elements that would integrate the abstract idea into a practical application.
Claim 12 recites the following additional elements:
wherein the training datasets each have a common format comprising a plurality of data blocks, wherein the plurality of data blocks includes a first data block containing training media for the respective training dataset, a second data block indicative of the use rights for the respective training dataset, a third data block indicative of media metadata corresponding to the training media in the first data block and optionally a fourth data block indicative of a unique user identifier for the respective training dataset
This limitation is directed towards a continuation of the insignificant extra solution activity of mere data gathering (see MPEP § 2106.05(g)).
Step 2B: Does the claim as a whole amount to significantly more than the judicial exception?
No, the claim as a whole does not amount to significantly more than the judicial exception. All elements of the claim, viewed individually or wholistically, do not provide an inventive concept or otherwise significantly more than the abstract idea itself.
Claim 12 recites the following additional elements:
wherein the training datasets each have a common format comprising a plurality of data blocks, wherein the plurality of data blocks includes a first data block containing training media for the respective training dataset, a second data block indicative of the use rights for the respective training dataset, a third data block indicative of media metadata corresponding to the training media in the first data block and optionally a fourth data block indicative of a unique user identifier for the respective training dataset
This limitation is directed towards a continuation of the insignificant extra solution activity of mere data gathering (see MPEP § 2106.05(g)). This is a well understood, routine, conventional activity of transmitting data (see MPEP 2106.05(d) example i in computer functions). Noting data that is included in the input does not integrate into a practical application.
In regards to Claim 14:
Step 2A Prong 2: Does the claim recite additional elements that integrate the exception into a practical application of the exception?
No, the application does not recite any additional elements that would integrate the abstract idea into a practical application.
Claim 14 recites the following additional elements:
wherein the media metadata corresponding to the training media in the first data block comprises a list of objects depicted in the training media
This limitation is directed towards a continuation of the insignificant extra solution activity of mere data gathering (see MPEP § 2106.05(g)).
Step 2B: Does the claim as a whole amount to significantly more than the judicial exception?
No, the claim as a whole does not amount to significantly more than the judicial exception. All elements of the claim, viewed individually or wholistically, do not provide an inventive concept or otherwise significantly more than the abstract idea itself.
Claim 14 recites the following additional elements:
wherein the media metadata corresponding to the training media in the first data block comprises a list of objects depicted in the training media
This limitation is directed towards a continuation of the insignificant extra solution activity of mere data gathering (see MPEP § 2106.05(g)). This is a well understood, routine, conventional activity of transmitting data (see MPEP 2106.05(d) example i in computer functions). Noting data that is included in the input does not integrate into a practical application.
In regards to Claim 15:
Step 2A Prong 2: Does the claim recite additional elements that integrate the exception into a practical application of the exception?
No, the application does not recite any additional elements that would integrate the abstract idea into a practical application.
Claim 15 recites the following additional elements:
wherein the media metadata corresponding to the training media in the first data block is further indicative of whether the training media was derived from real-world imaging or derived synthetically
This limitation is directed towards a continuation of the insignificant extra solution activity of mere data gathering (see MPEP § 2106.05(g)).
Step 2B: Does the claim as a whole amount to significantly more than the judicial exception?
No, the claim as a whole does not amount to significantly more than the judicial exception. All elements of the claim, viewed individually or wholistically, do not provide an inventive concept or otherwise significantly more than the abstract idea itself.
Claim 15 recites the following additional elements:
wherein the media metadata corresponding to the training media in the first data block is further indicative of whether the training media was derived from real-world imaging or derived synthetically
This limitation is directed towards a continuation of the insignificant extra solution activity of mere data gathering (see MPEP § 2106.05(g)). This is a well understood, routine, conventional activity of transmitting data (see MPEP 2106.05(d) example i in computer functions). Noting data that is included in the input does not integrate into a practical application.
In regards to Claim 17:
Step 2A Prong 1: Does the claim recite a law of nature, a natural phenomenon, or an abstract idea?
Yes, the claim does recite a(n) abstract idea.
Claim 17 recites the following abstract ideas:
wherein to identify a candidate set of training datasets matching the specifications, the method comprises comparing the specifications to the media metadata in the third data block of each of the training datasets, the candidate set of training datasets including those training datasets for which the media metadata in the third data block of the respective training datasets matches the specifications
This limitation is directed towards the abstract idea of a mental process, or a concept performed in the human mind, including observation, evaluation, judgement or opinion (see MPEP 2106.04(a)(2) subsection 3). Here the limitation is seen as evaluation and judgement.
In regards to Claim 20:
Step 2A Prong 1: Does the claim recite a law of nature, a natural phenomenon, or an abstract idea?
Yes, the claim does recite a(n) abstract idea.
Claim 20 recites the following abstract ideas:
wherein to identify a candidate set of training datasets matching the specifications, the method comprises comparing the user-specified application scope to the licensed application scope, the candidate set of training datasets including those training datasets for which the licensed application scope in the second data block of the respective training dataset includes the user-specified application scope
This limitation is directed towards the abstract idea of a mental process, or a concept performed in the human mind, including observation, evaluation, judgement or opinion (see MPEP 2106.04(a)(2) subsection 3). Here the limitation is seen as evaluation and judgement.
Step 2A Prong 2: Does the claim recite additional elements that integrate the exception into a practical application of the exception?
No, the application does not recite any additional elements that would integrate the abstract idea into a practical application.
Claim 20 recites the following additional elements:
wherein the specifications comprise a user-specified application scope, wherein the use rights in the second data block comprise information related to a licensed application scope for the respective training dataset
This limitation is directed towards a continuation of the insignificant extra solution activity of mere data gathering (see MPEP § 2106.05(g)).
Step 2B: Does the claim as a whole amount to significantly more than the judicial exception?
No, the claim as a whole does not amount to significantly more than the judicial exception. All elements of the claim, viewed individually or wholistically, do not provide an inventive concept or otherwise significantly more than the abstract idea itself.
Claim 20 recites the following additional elements:
wherein the specifications comprise a user-specified application scope, wherein the use rights in the second data block comprise information related to a licensed application scope for the respective training dataset
This limitation is directed towards a continuation of the insignificant extra solution activity of mere data gathering (see MPEP § 2106.05(g)). This is a well understood, routine, conventional activity of transmitting data (see MPEP 2106.05(d) example i in computer functions). Noting data that is included in the input does not integrate into a practical application.
In regards to Claim 21:
Step 2A Prong 1: Does the claim recite a law of nature, a natural phenomenon, or an abstract idea?
Yes, the claim does recite a(n) abstract idea.
Claim <X> recites the following abstract ideas:
wherein to identify a candidate set of training datasets matching the specifications, the method further comprises comparing the user-specified list of objects to the list of objects in the third data block of each of the training datasets, the candidate set of training datasets excluding those training datasets for which the list of objects in the third data block of the respective training dataset does not contain any of the objects in the user-specified list of objects
This limitation is directed towards the abstract idea of a mental process, or a concept performed in the human mind, including observation, evaluation, judgement or opinion (see MPEP 2106.04(a)(2) subsection 3). Here the limitation is seen as evaluation and judgement.
Step 2A Prong 2: Does the claim recite additional elements that integrate the exception into a practical application of the exception?
No, the application does not recite any additional elements that would integrate the abstract idea into a practical application.
Claim 21 recites the following additional elements:
wherein the specifications comprise a user- specified list of objects, wherein the media metadata corresponding to the training media in the first data block comprises information related to a list of objects depicted in the training media in the first data block
This limitation is directed towards a continuation of the insignificant extra solution activity of mere data gathering (see MPEP § 2106.05(g)).
Step 2B: Does the claim as a whole amount to significantly more than the judicial exception?
No, the claim as a whole does not amount to significantly more than the judicial exception. All elements of the claim, viewed individually or wholistically, do not provide an inventive concept or otherwise significantly more than the abstract idea itself.
Claim 21 recites the following additional elements:
wherein the specifications comprise a user- specified list of objects, wherein the media metadata corresponding to the training media in the first data block comprises information related to a list of objects depicted in the training media in the first data block
This limitation is directed towards a continuation of the insignificant extra solution activity of mere data gathering (see MPEP § 2106.05(g)). This is a well understood, routine, conventional activity of transmitting data (see MPEP 2106.05(d) example i in computer functions). Noting data that is included in the input does not integrate into a practical application.
In regards to Claim 22:
Step 2A Prong 2: Does the claim recite additional elements that integrate the exception into a practical application of the exception?
No, the application does not recite any additional elements that would integrate the abstract idea into a practical application.
Claim 22 recites the following additional elements:
wherein the authorizing comprises creating the subset of the candidate set of training datasets, the subset including those training datasets in the candidate set for which the use rights in the second data block of the respective training datasets include the proposed usage of the Al training data
At a high level of generality, this is an activity of using an element as an “apply it” use (see MPEP 2106.05(f)).
Step 2B: Does the claim as a whole amount to significantly more than the judicial exception?
No, the claim as a whole does not amount to significantly more than the judicial exception. All elements of the claim, viewed individually or wholistically, do not provide an inventive concept or otherwise significantly more than the abstract idea itself.
Claim 22 recites the following additional elements:
wherein the authorizing comprises creating the subset of the candidate set of training datasets, the subset including those training datasets in the candidate set for which the use rights in the second data block of the respective training datasets include the proposed usage of the Al training data
At a high level of generality, this is an activity of using an element as an “apply it” use (see MPEP 2106.05(f)). At said high level of generality, a generic recitation of “apply” or equivalent does not incorporate the abstract idea into a practical invention and is seen as a variation of the phrase “apply it”. Noting the identification is applied or utilized as a filter does not incorporate into a practical application.
In regards to Claim 23:
Step 2A Prong 2: Does the claim recite additional elements that integrate the exception into a practical application of the exception?
No, the application does not recite any additional elements that would integrate the abstract idea into a practical application.
Claim 23 recites the following additional elements:
wherein the proposed usage of the Al training data comprises an identifier of the Al platform, wherein the use rights in the second data block of each of the training datasets identifies at least one licensed Al platform
This limitation is directed towards a continuation of the insignificant extra solution activity of mere data gathering (see MPEP § 2106.05(g)).
wherein the authorizing comprises creating the subset of the candidate set of training datasets, the subset including those training datasets in the candidate set for which the at least one licensed Al platform identified by the second data block of the respective training dataset includes the identifier of the Al platform
At a high level of generality, this is an activity of using an element as an “apply it” use (see MPEP 2106.05(f)).
Step 2B: Does the claim as a whole amount to significantly more than the judicial exception?
No, the claim as a whole does not amount to significantly more than the judicial exception. All elements of the claim, viewed individually or wholistically, do not provide an inventive concept or otherwise significantly more than the abstract idea itself.
Claim 23 recites the following additional elements:
wherein the proposed usage of the Al training data comprises an identifier of the Al platform, wherein the use rights in the second data block of each of the training datasets identifies at least one licensed Al platform
This limitation is directed towards a continuation of the insignificant extra solution activity of mere data gathering (see MPEP § 2106.05(g)). This is a well understood, routine, conventional activity of transmitting data (see MPEP 2106.05(d) example i in computer functions). Noting data that is included in the input does not integrate into a practical application.
wherein the authorizing comprises creating the subset of the candidate set of training datasets, the subset including those training datasets in the candidate set for which the at least one licensed Al platform identified by the second data block of the respective training dataset includes the identifier of the Al platform
At a high level of generality, this is an activity of using an element as an “apply it” use (see MPEP 2106.05(f)). At said high level of generality, a generic recitation of “apply” or equivalent does not incorporate the abstract idea into a practical invention and is seen as a variation of the phrase “apply it”. Noting the identification is applied or utilized as a filter does not incorporate into a practical application.
In regards to Claim 24:
Step 2A Prong 2: Does the claim recite additional elements that integrate the exception into a practical application of the exception?
No, the application does not recite any additional elements that would integrate the abstract idea into a practical application.
Claim 24 recites the following additional elements:
wherein the proposed usage of the Al training data comprises a proposed geographic region, wherein the use rights in the second data block of each of the training datasets further comprises a licensed geographic region
This limitation is directed towards a continuation of the insignificant extra solution activity of mere data gathering (see MPEP § 2106.05(g)).
wherein the authorizing comprises creating the subset of the candidate set of training datasets, the subset including those training datasets in the candidate set for which the licensed geographic region in the second data block of the respective training dataset includes the proposed geographic region
At a high level of generality, this is an activity of using an element as an “apply it” use (see MPEP 2106.05(f)).
Step 2B: Does the claim as a whole amount to significantly more than the judicial exception?
No, the claim as a whole does not amount to significantly more than the judicial exception. All elements of the claim, viewed individually or wholistically, do not provide an inventive concept or otherwise significantly more than the abstract idea itself.
Claim 24 recites the following additional elements:
wherein the proposed usage of the Al training data comprises a proposed geographic region, wherein the use rights in the second data block of each of the training datasets further comprises a licensed geographic region
This limitation is directed towards a continuation of the insignificant extra solution activity of mere data gathering (see MPEP § 2106.05(g)). This is a well understood, routine, conventional activity of transmitting data (see MPEP 2106.05(d) example i in computer functions). Noting data that is included in the input does not integrate into a practical application.
wherein the authorizing comprises creating the subset of the candidate set of training datasets, the subset including those training datasets in the candidate set for which the licensed geographic region in the second data block of the respective training dataset includes the proposed geographic region
At a high level of generality, this is an activity of using an element as an “apply it” use (see MPEP 2106.05(f)). At said high level of generality, a generic recitation of “apply” or equivalent does not incorporate the abstract idea into a practical invention and is seen as a variation of the phrase “apply it”. Noting the identification is applied or utilized as a filter does not incorporate into a practical application.
In regards to Claim 34:
This claim recites limitations containing analogous subject matter to claim 1.
In regards to Claim 35:
This claim recites limitations containing analogous subject matter to claim 1. The introduction of modules does not change the 101 interpretation, as the modules perform the aspects as recited from claim 1.
Claim 35 is interpreted as not being directed towards software per se, as while claim 35 does not explicitly recite hardware such as memory, processor, or other, the specification paragraph 39 notes figure 27 is a block diagram of a computing system and figure 27 shows memory, processor, and more. Paragraph 2 of the specification indicates “The present disclosure relates to the management of data components used to train Artificial Intelligence (Al) platforms”, which indicates the management server parts of a computing system in paragraph 39 performs the limitations noted in claim 35.
A computer system is deemed to be akin to a computing system as paragraph 198 of the specification notes [Current Specification 0198] :“It should be appreciated that throughout the specification, discussions utilizing terms such as "processing", "computing", "calculating", "determining", "analyzing" or the like, can refer to the action and/or processes of a computer or computing system, or similar electronic computing device, that manipulate and/or transform data represented as physical, such as electronic, quantities into other data similarly represented as physical quantities.”
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.
The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or nonobviousness.
Total claims rejected under 35 U.S.C. 103 are 1-3, 5, 8-12, 14, 17, 20-24, 34.
Claims 1-3, 6, 8-12, 14, 17, 20, 22-24, 34, 35 is/are rejected under 35 U.S.C. 103 as being unpatentable over Weast and Neal.
Claim 1:
Weast teaches:
Below teaches all of the relevant elements for operating in a computer:
[Weast 0074]: “Example computer system 900 includes at least one processor 902 (e.g., a central processing unit (CPU) [processor], a graphics processing unit (GPU) or both, processor cores, compute nodes, etc.), a main memory [memory] 904 and a static memory 906, which communicate with each other via a link 908 (e.g., bus). The computer system 900 may further include a video display unit 910, an alphanumeric input device 912 (e.g., a keyboard), and a user interface (UI) [user interface] navigation device 914 (e.g., a mouse). In one embodiment, the video display unit 910, input device 912 and UI navigation device 914 are incorporated into a touch screen display. The computer system 900 may additionally include a storage device 916 (e.g., a drive unit), a signal generation device 918 (e.g., a speaker), a network interface device 920, and one or more sensors (not shown), such as a global positioning system (GPS) sensor, compass, accelerometer, gyrometer, magnetometer, or other sensor.”
[Weast 0076]: “The term “machine-readable medium” [computer readable medium] shall also be taken to include any tangible medium that is capable of storing, encoding or carrying instructions for execution by the machine and that cause the machine to perform any one or more of the methodologies of the present disclosure or that is capable of storing, encoding or carrying data structures utilized by or associated with such instructions.”
A method of operating a computer to build a customized training dataset for training an artificial intelligence (Al) platform, the method comprising
consulting a database of training datasets, each associated with use rights, to identify a candidate set of training datasets matching the specifications
and authorizing release of a subset of the training datasets in the candidate set of training datasets based on the data indicative of the proposed usage of the Al training data and the use rights associated with the training data sets in the candidate set of training datasets
[Weast 0017]: “In example embodiments, various systems and methods for customizing training data for an AI or ML model are disclosed [A method of operating a computer to build a customized training dataset for training an artificial intelligence (Al) platform, the method comprising]. A set of data is identified [consulting a database of training datasets, each associated with use rights, to identify a candidate set of training datasets matching the specifications where use rights is indicated by the other parts of the quote noting the metadata is used for the identification where jurisdiction of something is accounted for] from a plurality of sets of data used to train the AI or ML model. The set of data is identified based on a set of metadata associated with the set of data indicating an association between the set of data and a jurisdiction of a digital services tax (DST). Based on the identifying, the plurality of sets of data is modified by removing the set of data, reducing the set of data, or reducing weights or influence values associated with the set of data [and authorizing release of a subset of the training datasets in the candidate set of training datasets based on the data indicative of the proposed usage of the Al training data and the use rights associated with the training data sets in the candidate set of training datasets as removing data that is not approved for the usage or as a result of jurisdiction/policies means only an authorized subset is output or used for training which is shown by a model using the data as training data in this paragraph]. The AI or ML model is retrained based on the modified plurality of sets of data. The retrained AI or ML model is provided for deployment in an individual autonomous agent. In example embodiments, the identifying of the set of metadata is based on a change being detected in the DST. In example embodiments, a traceable receipt or certificate travels with the retrained model for verification of the modification and/or the set of data (e.g., with respect to origins of each data point in the set of data or the plurality of sets of data).”
Support for removal based on jurisdiction or policies is shown in [Weast 0016]: “Unlike existing solutions, which require starting from scratch with data collection all over again for a particular application (e.g., because their training data sets do not have traceability based on where the data was gathered for purposes of ensuring any model trained with that data is restricted to use only in the region where the training data was originally collected), the disclosed embodiments are configured to differentiate between unregulated vs. regulated training material, allowing models to be deployed at scale without triggering potentially costly DSTs. In example embodiments, an audit trail is generated (e.g., via a block chain) for a set of data. In example embodiments, the audit trail may be used to indicate to a government whether an entity owes a DST to the government for use of the set of data, such as for use in training of AI/ML models. In example embodiments, a probability that a particular set of data will become regulated or unregulated is calculated. Based on the probability, the set of data may be flagged for removal or reduction in use.”
Distribution rights on information is supported in [Weast 0028]: “In example embodiments, the content being generated can be attached to a DRM license managed by each jurisdiction (e.g., including a government or corporation, such as an original design manufacturer (ODM)) to enforce that specific data with specific tags are used in the right place.”
Weast does not explicitly teach:
obtaining (i) specifications of training data for training the Al platform and (ii) data indicative of a proposed usage of the training data
Weast indicates aspects of the above limitation, such as usage as part of the data ([Weast 0065]: “Each data point and/or data set may be associated with metadata that allows for traceability of source and/or usage of the data, as described herein.”), but input such as via UI is more completely taught via Neal.
Neal teaches:
obtaining (i) specifications of training data for training the Al platform and (ii) data indicative of a proposed usage of the training data
[Neal 0110]: “Referring back to FIG. 11, the right portion side portion of GUI instance 1100 is a data usage selection GUI 1130 (labeled “how are you using the data?”), which is used to specify how the data is being used in the global context 722g and/or local context 722b [obtaining (i) specifications of training data for training the Al platform and (ii) data indicative of a proposed usage of the training data where further support is shown in Figure 8 below]. The GUI 1130 includes various GCEs 1131 for selecting the data usage(s) for this org context 722. FIG. 14 shows a GUI instance 1400 including a GUI element 1430 for adding and/or removing different data usage taxonomies. In this example, the GUI element 1430 includes prepopulated taxonomies that have a number of different data uses. As can be seen in FIG. 14, all of the taxonomies are hierarchical. The user can also make changes to how the taxonomies are listed and arranged within the hierarchy to fit their particular org 120.”
[Neal Figure 8]:
PNG
media_image1.png
916
916
media_image1.png
Greyscale
One of ordinary skill in the art, prior to the effective filing date, would have been motivated to combine Weast and Neal. Weast and Neal are in the same field of endeavor of information handling and machine learning. One of ordinary skill in the art would be motivated to combine Weast and Neal for incorporating aspects of filtering information to ensure being compliant with privacy laws, regulations, or such ([Neal Abstract]: “The present disclosure describes an adaptive data privacy platform that facilitates compliance with privacy laws and regulations, and compliance with organizational requirements within an organizational context. Other embodiments and implementations may be described and/or claimed.”)
Claim 2:
The method defined in claim 1 is taught by Weast and Neal.
Neal teaches:
wherein the specifications include an application scope.
[Neal 0077]: “The set of hierarchical metadata scope tags [wherein the specifications include an application scope] 210c are used by the org modeler 255 (e.g., question wizard and/or other suitable GUI elements) to allow users related to the org to provide an org model. The org modeler 255 allows users to define or specify various data types, use cases, geographies and/or jurisdictions, and/or other conditions or parameters that are applicable to an org and/or how an org handles data. This org modeler 255 produces a global org context model 225 based on inputs provided to the org modeler 255. In some implementations, the global org context model 225 is the same or similar as the org context model 220. The global org context model 225 can be reduced and/or divided into a set of local org contexts 227 (also referred to as “local org context models 227”, “child org contexts 227”, and/or the like). The global org context model 260 can be broken into an arbitrary number of children org contexts 227 by reducing the number of applicable tags 210. The local org contexts 227 are filtered views of the org context model 220 (or global org context model 225), which is set up during the configuration (aspect 200). The local org contexts 227 represent the org's 120 components or units by one or more parameters and/or characteristic. The set of local org contexts 227 can be further filtered by data type(s), use case(s), geographies, and/or other parameters. Children org contexts 227 cannot extend beyond parent org context 225 boundaries. Each child org contexts 227 can also have one or more child org contexts 227. In some implementations, each child org context 227 has at least one geo (geography tag), one data type, and one use case to be valid.”
The motivation to combine with Neal is the same as in claim 1.
Claim 3:
The method defined in claim 2 is taught by Weast and Neal.
Neal teaches:
wherein the application scope includes at least one of (i) a global application; and (ii) a specific application.
[Neal 0077]: “The set of hierarchical metadata scope tags 210c are used by the org modeler 255 (e.g., question wizard and/or other suitable GUI elements) to allow users related to the org to provide an org model. The org modeler 255 allows users to define or specify various data types, use cases, geographies and/or jurisdictions, and/or other conditions or parameters that are applicable to an org and/or how an org handles data. This org modeler 255 produces a global org context model 225 based on inputs provided to the org modeler 255. In some implementations, the global org context model [wherein the application scope includes at least one of (i) a global application] 225 is the same or similar as the org context model 220. The global org context model 225 can be reduced and/or divided into a set of local org contexts [and (ii) a specific application] 227 (also referred to as “local org context models 227”, “child org contexts 227”, and/or the like). The global org context model 260 can be broken into an arbitrary number of children org contexts 227 by reducing the number of applicable tags 210. The local org contexts 227 are filtered views of the org context model 220 (or global org context model 225), which is set up during the configuration (aspect 200). The local org contexts 227 represent the org's 120 components or units by one or more parameters and/or characteristic. The set of local org contexts 227 can be further filtered by data type(s), use case(s), geographies, and/or other parameters. Children org contexts 227 cannot extend beyond parent org context 225 boundaries. Each child org contexts 227 can also have one or more child org contexts 227. In some implementations, each child org context 227 has at least one geo (geography tag), one data type, and one use case to be valid.”
The motivation to combine with Neal is the same as in claim 1.
Claim 6:
The method defined in claim 1 is taught by Weast and Neal.
Weast teaches:
wherein the data indicative of a proposed usage of the training data includes one or more of (i) an identifier of the AI platform and (ii) a proposed geographic region.
Support for the data being indicative of a usage and such is provided in [Weast 0065]: “Each data point and/or data set may be associated with metadata that allows for traceability of source and/or usage of the data, as described herein.”
[Weast 0036]: “FIG. 4 is an example of distribution of policy management 400 across different entities [wherein the data indicative of a proposed usage of the training data includes one or more of (i) an identifier of the AI platform as this shows entities or AI platforms are important in how the decision of distributing information as there are policies related to the entities or AI platforms]. As shown, a TEE in one or more privacy sensitive base stations makes policy decisions. A TEE in individual autonomous agent(s) enforces the policies via the TEE (e.g., for an array of sensors)”
Further support for limiting by agent/AI platform and location is indicated in [Weast 0035]: “A DST policy generation manager 312 is configured to implement a policy-based action if an individual autonomous agent cannot comply with a requested policy or enforce the requested policy. A geo-fencing manager 314 is configured to provide an attestation response token that includes DST content sharing polices in a geo-fenced restricted zone that is to be enforced via the TEE.”
Aspects for sensors for agents or vehicles is indicated in the above and below quotes, which are considered relevant to help provide indication of AI platform, as the specification paragraph 44 and the WO noted in paragraph 46 indicate AI agents or models attached to sensors in relation to AI platform. Weast provides a description of aspects of the sensor setup in paragraph 20.
[Weast 0028]: “In example embodiments, the content being generated can be attached to a DRM license managed by each jurisdiction (e.g., including a government or corporation, such as an original design manufacturer (ODM)) to enforce that specific data with specific tags are used in the right place [and (ii) a proposed geographic region as this indicates the requirement of checking the region for data usage]. In example embodiments, depending on a tax agreement with a country, the country may be banned from using certain data or an amount of use of that certain data by the country may be restricted. Alternatively, this could be done by using geographic based certificates that sign the data using a certificate issued by the controlling government.”
Neal additionally teaches:
and (ii) a proposed geographic region
Further support for geographic region as input is provided by Neal by figure 8 and the below paragraph:
[Neal 0077]: “The set of local org contexts 227 can be further filtered by data type(s), use case(s), geographies, and/or other parameters. Children org contexts 227 cannot extend beyond parent org context 225 boundaries. Each child org contexts 227 can also have one or more child org contexts 227. In some implementations, each child org context 227 has at least one geo (geography tag) [and (ii) a proposed geographic region], one data type, and one use case to be valid.”
The motivation to combine with Neal is the same as in claim 1.
Claim 8:
The method defined in claim 1 is taught by Weast and Neal.
Weast teaches:
wherein the authorizing comprises creating a subset of the candidate set of training datasets, the subset including those training datasets in the candidate set for which the use rights include the proposed usage of the AI training data.
[Weast 0017]: “In example embodiments, various systems and methods for customizing training data for an AI or ML model are disclosed. A set of data is identified from a plurality of sets of data used to train the AI or ML model. The set of data is identified based on a set of metadata associated with the set of data indicating an association between the set of data and a jurisdiction of a digital services tax (DST). Based on the identifying, the plurality of sets of data is modified by removing the set of data, reducing the set of data, or reducing weights or influence values associated with the set of data [wherein the authorizing comprises creating a subset of the candidate set of training datasets, the subset including those training datasets in the candidate set for which the use rights include the proposed usage of the AI training data as indicated by the teachings for similar limitation in claim 1, as claim 1 explains the removal of regulated material]. The AI or ML model is retrained based on the modified plurality of sets of data. The retrained AI or ML model is provided for deployment in an individual autonomous agent. In example embodiments, the identifying of the set of metadata is based on a change being detected in the DST. In example embodiments, a traceable receipt or certificate travels with the retrained model for verification of the modification and/or the set of data (e.g., with respect to origins of each data point in the set of data or the plurality of sets of data).”
As noted in claim 1, information having distribution or usage rights in known in Weast ([Weast 0028]: “In example embodiments, the content being generated can be attached to a DRM license managed by each jurisdiction (e.g., including a government or corporation, such as an original design manufacturer (ODM)) to enforce that specific data with specific tags are used in the right place.”)
Support for the data being indicative of a usage and such is provided in [Weast 0065]: “Each data point and/or data set may be associated with metadata that allows for traceability of source and/or usage of the data, as described herein.”
The premise of authorized access is shared in other information distribution platforms or systems, such as in Neal ([Neal 0048]: “Additionally, the requirements can specify (within various levels of granularity) particular users, and user groups, that are authorized to access particular data types, based on the org's hierarchical structure, and security and regulatory requirements.”)
Claim 9:
The method defined in claim 8 is taught by Weast and Neal.
Weast teaches:
wherein the subset of the candidate set of training datasets excludes any training datasets in the candidate set for which the use rights do not include the proposed usage of the AI training data.
[Weast 0017]: “In example embodiments, various systems and methods for customizing training data for an AI or ML model are disclosed. A set of data is identified from a plurality of sets of data used to train the AI or ML model. The set of data is identified based on a set of metadata associated with the set of data indicating an association between the set of data and a jurisdiction of a digital services tax (DST). Based on the identifying, the plurality of sets of data is modified by removing the set of data [wherein the subset of the candidate set of training datasets excludes any training datasets in the candidate set for which the use rights do not include the proposed usage of the AI training data], reducing the set of data, or reducing weights or influence values associated with the set of data. The AI or ML model is retrained based on the modified plurality of sets of data. The retrained AI or ML model is provided for deployment in an individual autonomous agent. In example embodiments, the identifying of the set of metadata is based on a change being detected in the DST. In example embodiments, a traceable receipt or certificate travels with the retrained model for verification of the modification and/or the set of data (e.g., with respect to origins of each data point in the set of data or the plurality of sets of data).”
As discussed in claim 8, the restriction of using allowed data and not using not allowed data is indicated by the filtering or removal of information based on the rights/permissions.
Claim 10:
The method defined in claim 1 is taught by Weast and Neal.
Neal teaches:
wherein the specifications and the data indicative of a proposed usage of the training data are obtained from a user, the method further comprising determining whether the user has a license to train the AI platform.
[Neal 0110]: “Referring back to FIG. 11, the right portion side portion of GUI instance 1100 is a data usage selection GUI 1130 (labeled “how are you using the data?”), which is used to specify how the data is being used in the global context 722g and/or local context 722b [wherein the specifications and the data indicative of a proposed usage of the training data are obtained from a user]. The GUI 1130 includes various GCEs 1131 for selecting the data usage(s) for this org context 722. FIG. 14 shows a GUI instance 1400 including a GUI element 1430 for adding and/or removing different data usage taxonomies. In this example, the GUI element 1430 includes prepopulated taxonomies that have a number of different data uses. As can be seen in FIG. 14, all of the taxonomies are hierarchical. The user can also make changes to how the taxonomies are listed and arranged within the hierarchy to fit their particular org 120.”
[Neal 0048]: “Additionally, the requirements can specify (within various levels of granularity) particular users [the method further comprising determining whether the user has a license to train the AI platform], and user groups, that are authorized to access particular data types, based on the org's hierarchical structure, and security and regulatory requirements.”
[Neal 0304]: “The term “legal metadata” at least in some embodiments refers to information about a creator of content or data item, copyright holder, licensing information, and/or other legalistic information.”
Motivation to combine with Neal is the same as claim 1.
Neal is utilized to teach claim 10, but relevant aspects from Weast are noted below:
Notes aspects of license and such requirements are relevant
[Weast 0026]: “In example embodiments, the autonomous vehicle(s) 102 communicate over the network 112 with a license infrastructure 114. In example embodiments, the license infrastructure 114 includes a license server 116 and a training content server 118. In example embodiments, the training content server(s) 118 are configured to train one or more AI or ML models for deployment in the autonomous vehicle(s) 102. In example embodiments, the license server(s) 116 are configured to identify any licensing or DST requirements associated with the data used by the training content server(s) 118 and to optimize the training data to minimize DSTs, as described in more detail below.”
Notes licensing for ensuring used only by correct recipients
[Weast 0028]: “In example embodiments, the content being generated can be attached to a DRM license managed by each jurisdiction (e.g., including a government or corporation, such as an original design manufacturer (ODM)) to enforce that specific data with specific tags are used in the right place.”
Claim 11:
The method defined in claim 10 is taught by Weast and Neal.
Neal teaches:
wherein at least some of the training datasets in the database are encrypted and wherein if the user is determined to have a license to train the AI platform, the consulting comprises a step of decrypting the at least some of the training datasets in the database.
[Neal 0137]: “Furthermore, one or more ADPP servers 145 may hash, digitally sign, and/or encrypt/decrypt data [wherein at least some of the training datasets in the database are encrypted and wherein if the user is determined to have a license to train the AI platform, the consulting comprises a step of decrypting the at least some of the training datasets in the database] using, for example, a cryptographic hash algorithm, such as a function in the Secure Hash Algorithm (SHA) 2 set of cryptographic hash algorithms (e.g., SHA-226, SHA-256, SHA-512, and the like)”
Neal, as discussed using paragraphs below, indicates the knowledge and use of cryptographic principles and information security. With the use of encryption and decryption in relation to data from the server handling the information (as indicated by the above quote from paragraph 137), the premise of securing information to only those with a license would be obvious to one of ordinary skill in the art, prior to the effective filing date, when combined with the teachings for limiting by license teachings from claim 10. In the field of endeavor of information handling or distribution, the premise of security for authorization to information or data is known (Neal 0048 notes the limiting to particular users or such for regulation or policy and Neal 0304 notes the legal requirements or policies related to information being stored in metadata for data). The paragraphs below from Neal indicate the use and motivation of security elements such as encryption which when practiced with information security would involve “mitigating information risks” as discussed in paragraph 0325 of Neal.
[Neal 0322]: “The term “cryptographic mechanism” at least in some embodiments refers to any cryptographic protocol and/or cryptographic algorithm. Additionally or alternatively, the term “cryptographic protocol” at least in some embodiments refers to a sequence of steps precisely specifying the actions required of two or more entities to achieve specific security objectives (e.g., cryptographic protocol for key agreement). Additionally or alternatively, the term “cryptographic algorithm” at least in some embodiments refers to an algorithm specifying the steps followed by a single entity to achieve specific security objectives (e.g., cryptographic algorithm for symmetric key encryption).”
[Neal 0325]: “The term “information security” or “InfoSec” at least in some embodiments refers to any practice, technique, and technology for protecting information by mitigating information risks and typically involves preventing or reducing the probability of unauthorized/inappropriate access to data, or the unlawful use, disclosure, disruption, deletion, corruption, modification, inspection, recording, or devaluation of information; and the information to be protected may take any form including electronic information, physical or tangible (e.g., computer-readable media storing information, paperwork, and the like), or intangible (e.g., knowledge, intellectual property assets, and the like).”
Support from primary ref of using encryption [Weast 0048]: “FIG. 7 depicts a method 700 of a DST flow with block chain support. At 1, raw harvested data, geo-tags, and/or provenance data is encrypted and signed via a device-specific key.”
Claim 12:
The method defined in claim 1 is taught by Weast and Neal.
Weast teaches:
wherein the training datasets each have a common format comprising a plurality of data blocks, wherein the plurality of data blocks includes a first data block containing training media for the respective training dataset, a second data block indicative of the use rights for the respective training dataset, a third data block indicative of media metadata corresponding to the training media in the first data block and optionally a fourth data block indicative of a unique user identifier for the respective training dataset.
[Weast 0028]: “During the harvesting of the data [wherein the training datasets each have a common format comprising a plurality of data blocks, wherein the plurality of data blocks includes a first data block containing training media for the respective training dataset], the data is tagged with metadata [a third data block indicative of media metadata corresponding to the training media in the first data block], such as one or more of a geo-location tag, date, time, day of week, a number of humans nearby, data pertaining to a type of a location of the harvesting, and so on. In example embodiments, the metadata may include any data pertaining to the location of the harvesting of the data that is identified as being relevant to whether a DST will be applied and/or an amount of the DST. For example, day of week and time metadata may be relevant to surge pricing of DSTs, such as when a DST is applied Monday-Friday from 9-5 pm (e.g., based on a judgment that weekday data is more valuable than weekend data). In example embodiments, the metadata is stored with and/or associated with the harvested data. In example embodiments, the geo-location tag or other location-related metadata may define a geographical area (e.g., that is delimited with a set of points). In example embodiments, the location-related metadata may be identified as corresponding to a region or a country. In example embodiments, the location-related metadata may define a type of the location (e.g., urban versus countryside). In example embodiments, such location-related metadata is added to every piece of training data so that there is traceability from data collection for training through deployment of that model for commercial purposes. In example embodiments, every tag and training data is signed by the entity that generates it [and optionally a fourth data block indicative of a unique user identifier for the respective training dataset (a possible interpretation that user in this case refers to what the data is from or related to)]. In this way, the entity that generated the data may be validated and it may be validated that the tag has not been tampered with. In example embodiments, the content being generated can be attached to a DRM license [a second data block indicative of the use rights for the respective training dataset] managed by each jurisdiction (e.g., including a government or corporation, such as an original design manufacturer (ODM)) to enforce that specific data with specific tags are used in the right place. In example embodiments, depending on a tax agreement with a country, the country may be banned from using certain data or an amount of use of that certain data by the country may be restricted. Alternatively, this could be done by using geographic based certificates that sign the data using a certificate issued by the controlling government.”
Context for showing the above data collection is related to training data is given in the paragraph before indicating the data collection is for training data, thus the elements taught above indicate training data containing the requiring information related to training data ([Weast 0027]: “At operation 202, data used for training one or more AI or ML models is harvested. For example, the data is harvested from a secure storage in one or more of the vehicle(s) 102 during operation of those vehicles in a specific government jurisdiction. In example embodiments, the data may include data collected from sensors of the vehicle during operation of the vehicle.”).
Claim 14:
The method defined in claim 12 is taught by Weast and Neal.
Weast teaches:
wherein the media metadata corresponding to the training media in the first data block comprises a list of objects depicted in the training media.
[Weast 0055]: “In example embodiments, machine learning for fine-grained data tagging is implemented. The training data is tagged with scene description [wherein the media metadata corresponding to the training media in the first data block comprises a list of objects depicted in the training media] to show multiple features to help the decision on regulated versus un-regulated data. Examples: (1) Data collected on a highway in country1/region1 that is a common highway with country2/region2 can be regulated data that does not need taxation in country2/region2; (2) Data relative to safe driving (e.g., data including traffic signs) may not be subject to taxation; (3) Data pertaining to pedestrians may be regulated or unregulated in some regions based on privacy laws.”
Claim 17:
The method defined in claim 1 is taught by Weast and Neal.
Weast teaches:
wherein to identify a candidate set of training datasets matching the specifications, the method comprises comparing the specifications to the media metadata in the third data block of each of the training datasets, the candidate set of training datasets including those training datasets for which the media metadata in the third data block of the respective training datasets matches the specifications.
[Weast 0017]: “In example embodiments, various systems and methods for customizing training data for an AI or ML model are disclosed. A set of data is identified from a plurality of sets of data used to train the AI or ML model. The set of data is identified based on a set of metadata associated with the set of data [wherein to identify a candidate set of training datasets matching the specifications, the method comprises comparing the specifications to the media metadata in the third data block of each of the training datasets, the candidate set of training datasets including those training datasets for which the media metadata in the third data block of the respective training datasets matches the specifications] indicating an association between the set of data and a jurisdiction of a digital services tax (DST). Based on the identifying, the plurality of sets of data is modified by removing the set of data, reducing the set of data, or reducing weights or influence values associated with the set of data. The AI or ML model is retrained based on the modified plurality of sets of data. The retrained AI or ML model is provided for deployment in an individual autonomous agent. In example embodiments, the identifying of the set of metadata is based on a change being detected in the DST. In example embodiments, a traceable receipt or certificate travels with the retrained model for verification of the modification and/or the set of data (e.g., with respect to origins of each data point in the set of data or the plurality of sets of data).”
Alternative teaching of similar ideas from Neal
[Neal 0053]: “The ADPP 140 includes metadata tagging, schemas, and filtering aspects that interact with one another. The metadata tagging, schemas, and filtering mechanisms link various components in the ADPP 140 to provide a user 105, 120 with successful privacy program. Unlike existing approaches where many orgs 120 apply the same “one-size-fits-none” framework to their privacy program, users of the ADPP 140 create a model of their org's 120 structure and use(s) of personal information using metadata tags (e.g., metadata tags 210a, 210b, and 210c of FIG. 2). The ADPP 140 uses the metadata tags as a multi-dimensional filter [wherein to identify a candidate set of training datasets matching the specifications, the method comprises comparing the specifications to the media metadata in the third data block of each of the training datasets, the candidate set of training datasets including those training datasets for which the media metadata in the third data block of the respective training datasets matches the specifications] against an extensive content catalog of privacy laws, regulations, and other requirements (e.g., content catalog 240 of FIG. 2). In this way, an org 120 is defined as a hierarchical set of tags (e.g., metadata hierarchical scope tags 210c of FIG. 2) that are algorithmically compared against the tags associated with each requirement within the content catalog. When a tag in the org hierarchy matches a tag in the content catalog, a requirement associated with the tag in the content catalog is pulled into the org's 120 privacy program. Requirements could have several tags. For example, a tag and/or requirement may be “applies to employees in Belgium” or “applies to sending email in the USA” which allows the ADPP 140 to apply only parts of a law or framework that are applicable—rather than requiring a user to analyze a framework themselves to determine which elements are applicable to their org 120.”
The motivation to combine with Neal is the same as in claim 1.
Claim 20:
The method defined in claim 10 is taught by Weast and Neal.
Neal teaches:
wherein the specifications comprise a user-specified application scope, wherein the use rights in the second data block comprise information related to a licensed application scope for the respective training dataset, wherein to identify a candidate set of training datasets matching the specifications, the method comprises comparing the user-specified application scope to the licensed application scope, the candidate set of training datasets including those training datasets for which the licensed application scope in the second data block of the respective training dataset includes the user-specified application scope.
[Neal 0077]: “The set of hierarchical metadata scope tags [wherein the specifications comprise a user-specified application scope] 210c are used by the org modeler 255 (e.g., question wizard and/or other suitable GUI elements) to allow users related to the org to provide an org model. The org modeler 255 allows users to define or specify various data types, use cases, geographies and/or jurisdictions, and/or other conditions or parameters that are applicable to an org and/or how an org handles data. This org modeler 255 produces a global org context model 225 based on inputs provided to the org modeler 255. In some implementations, the global org context model 225 is the same or similar as the org context model 220. The global org context model 225 can be reduced and/or divided into a set of local org contexts 227 (also referred to as “local org context models 227”, “child org contexts 227”, and/or the like). The global org context model 260 can be broken into an arbitrary number of children org contexts 227 by reducing the number of applicable tags 210. The local org contexts 227 are filtered views of the org context model 220 (or global org context model 225), which is set up during the configuration (aspect 200). The local org contexts 227 represent the org's 120 components or units by one or more parameters and/or characteristic. The set of local org contexts 227 can be further filtered by data type(s), use case(s), geographies, and/or other parameters. Children org contexts 227 cannot extend beyond parent org context 225 boundaries. Each child org contexts 227 can also have one or more child org contexts 227. In some implementations, each child org context 227 has at least one geo (geography tag), one data type, and one use case to be valid.”
Neal also supports the idea of licensing being attached to data as well ([Neal 0304]: “The term “legal metadata” at least in some embodiments refers to information about a creator of content or data item, copyright holder, licensing information [Alternative teaching for wherein the use rights in the second data block comprise information related to a licensed application scope for the respective training dataset], and/or other legalistic information.”).
In regards to the combination of Weast and Neal:
Weast notes aspects of license and such requirements are checked in [Weast 0026]: “In example embodiments, the autonomous vehicle(s) 102 communicate over the network 112 with a license infrastructure 114. In example embodiments, the license infrastructure 114 includes a license server 116 and a training content server 118. In example embodiments, the training content server(s) 118 are configured to train one or more AI or ML models for deployment in the autonomous vehicle(s) 102. In example embodiments, the license server(s) 116 are configured to identify any licensing or DST requirements associated with the data used by the training content server(s) 118 and to optimize the training data to minimize DSTs, as described in more detail below.”
Weast notes licensing for ensuring used only by correct recipients [Weast 0028]: “In example embodiments, the content being generated can be attached to a DRM license [wherein the use rights in the second data block comprise information related to a licensed application scope for the respective training dataset] managed by each jurisdiction (e.g., including a government or corporation, such as an original design manufacturer (ODM)) to enforce that specific data with specific tags are used in the right place.”
The teachings of limiting/filtering by licensing (as noted above by Weast and Neal) when combined with the teaching of filtering of scope (as noted above by Neal) results in the above limitation of providing information within scope that fits the licensing is taught as the result of the comparison in the limitation is the result of output of multiple filters [wherein to identify a candidate set of training datasets matching the specifications, the method comprises comparing the user-specified application scope to the licensed application scope, the candidate set of training datasets including those training datasets for which the licensed application scope in the second data block of the respective training dataset includes the user-specified application scope].
Claim 22:
The method defined in claim 12 is taught by Weast and Neal.
Weast teaches:
wherein the authorizing comprises creating the subset of the candidate set of training datasets, the subset including those training datasets in the candidate set for which the use rights in the second data block of the respective training datasets include the proposed usage of the Al training data.
Notes aspects of license and such requirements are relevant [Weast 0026]: “In example embodiments, the autonomous vehicle(s) 102 communicate over the network 112 with a license infrastructure 114. In example embodiments, the license infrastructure 114 includes a license server 116 and a training content server 118. In example embodiments, the training content server(s) 118 are configured to train one or more AI or ML models for deployment in the autonomous vehicle(s) 102. In example embodiments, the license server(s) 116 are configured to identify any licensing or DST requirements [wherein the authorizing comprises creating the subset of the candidate set of training datasets, the subset including those training datasets in the candidate set for which the use rights in the second data block of the respective training datasets include the proposed usage of the Al training data where further support for rights related to the data is shown in paragraph 28 from Weast and a subset is made as the removing of information not fitting the requirements creates a subset (as noted in claim 1)] associated with the data used by the training content server(s) 118 and to optimize the training data to minimize DSTs, as described in more detail below.”
Notes licensing for ensuring used only by correct recipients [Weast 0028]: “In example embodiments, the content being generated can be attached to a DRM license managed by each jurisdiction (e.g., including a government or corporation, such as an original design manufacturer (ODM)) to enforce that specific data with specific tags are used in the right place.”
Neal provides additional teachings of limiting by usage, such as require a usage selection for the data in paragraph 110, and Neal indicates the usage is important for things like managing legal obligations, privacy, and such in paragraphs such as paragraph 35.
[Neal 0110]: “Referring back to FIG. 11, the right portion side portion of GUI instance 1100 is a data usage selection GUI 1130 (labeled “how are you using the data?”), which is used to specify how the data is being used in the global context 722g and/or local context 722b. The GUI 1130 includes various GCEs 1131 for selecting the data usage(s) for this org context 722. FIG. 14 shows a GUI instance 1400 including a GUI element 1430 for adding and/or removing different data usage taxonomies. In this example, the GUI element 1430 includes prepopulated taxonomies that have a number of different data uses. As can be seen in FIG. 14, all of the taxonomies are hierarchical. The user can also make changes to how the taxonomies are listed and arranged within the hierarchy to fit their particular org 120.”
[Neal 0035]: “The data privacy model may include or describe an org's 120 legal obligations affecting the personal information lifecycle, contractual obligations that restrict what an org 120 can do with personal information, the org's 120 strategic goals to the extent they are reliant on personal information, the org's 120 standard of ethical usage of personal information including transparency and fairness, and customer (or subscriber) satisfaction considerations, perspectives, and expectations. Data privacy models begin with the premise that legal obligations are not equivalent to privacy protections. In addition, contractual obligations may restrict what an org 120 can do with personal information.”
The motivation to combine with Neal is the same as in claim 1.
Claim 23:
The method defined in claim 22 is taught by Weast and Neal.
Weast teaches:
wherein the use rights in the second data block of each of the training datasets identifies at least one licensed Al platform
[Weast 0028]: “In example embodiments, the content being generated can be attached to a DRM license [wherein the use rights in the second data block of each of the training datasets identifies at least one licensed Al platform as this indicates the data for training sets has licenses associated with the training data or sets, where aspects of support related to AI platform are given in quotes below] managed by each jurisdiction (e.g., including a government or corporation, such as an original design manufacturer (ODM)) to enforce that specific data with specific tags are used in the right place.”
The below quotes from Weast indicate the premise of AI platforms, as indicated in the teachings of claim 6. Weast 0036 indicates the requirement of policies and such being enforced for an AI platform/agent.
[Weast 0064]: “In one example embodiment, the features 802 may be of different types and may include one or more of vehicle sensor array data, vehicle driving commands, or context data (e.g., a type of location, such as an intersection, that is inferred from sensor data, such as GPS coordinates; vehicle driving policy data, or other data inferred from the type of the location, time of day, or other metadata relevant to a context of an operation of the vehicle, such as risk data associated with operating the vehicle, or metadata pertaining to whether a DST is applicable or an amount of a DST with respect to particular data points in a data set).”
[Weast 0036]: “FIG. 4 is an example of distribution of policy management 400 across different entities. As shown, a TEE in one or more privacy sensitive base stations makes policy decisions. A TEE in individual autonomous agent(s) enforces the policies via the TEE (e.g., for an array of sensors)”
Neal teaches:
wherein the proposed usage of the Al training data comprises an identifier of the Al platform,
[Neal 0110]: “Referring back to FIG. 11, the right portion side portion of GUI instance 1100 is a data usage selection [wherein the proposed usage of the Al training data comprises an identifier of the Al platform where this is teaching the premise of input usage, where the usage being able to contain AI platform is indicated later in the existence of such data by Weast] GUI 1130 (labeled “how are you using the data?”), which is used to specify how the data is being used in the global context 722g and/or local context 722b. The GUI 1130 includes various GCEs 1131 for selecting the data usage(s) for this org context 722. FIG. 14 shows a GUI instance 1400 including a GUI element 1430 for adding and/or removing different data usage taxonomies. In this example, the GUI element 1430 includes prepopulated taxonomies that have a number of different data uses. As can be seen in FIG. 14, all of the taxonomies are hierarchical. The user can also make changes to how the taxonomies are listed and arranged within the hierarchy to fit their particular org 120.”
The motivation to combine with Neal is the same as claim 1.
Limiting to particular users, groups, etc is also supported by the teachings in Neal.
[Neal 0048]: “Additionally, the requirements can specify (within various levels of granularity) particular users, and user groups, that are authorized to access particular data types, based on the org's hierarchical structure, and security and regulatory requirements.”
The teachings from the combination of Weast and Neal:
wherein the authorizing comprises creating the subset of the candidate set of training datasets, the subset including those training datasets in the candidate set for which the at least one licensed Al platform identified by the second data block of the respective training dataset includes the identifier of the Al platform
The authorization of information or the filtering to allowed information is taught in claim 1. Indications of licensing being a reason to limit or filter by is provided by both Weast and Neal as indicated above. Weast in paragraph 36 indicates policies being done in relation to agents (vehicle with array of sensors) which is policies being done in relation to a AI platform. When combining the premises with the known ability to search or filter, which is indicated in both West ([Weast 0017]: “In example embodiments, various systems and methods for customizing training data for an AI or ML model are disclosed. A set of data is identified from a plurality of sets of data used to train the AI or ML model. The set of data is identified based on a set of metadata associated with the set of data indicating an association between the set of data and a jurisdiction of a digital services tax (DST). Based on the identifying, the plurality of sets of data is modified by removing the set of data, reducing the set of data, or reducing weights or influence values associated with the set of data.”) and Neal ([Neal 0048]: “Additionally, the requirements can specify (within various levels of granularity) particular users, and user groups, that are authorized to access particular data types, based on the org's hierarchical structure, and security and regulatory requirements.”), with the aspects of licensing being a reason to filter (as previously indicated), then the filtering for AI platform in relation to licensing is an obvious task one of ordinary skill in the art would perform. One of ordinary skill in the art would be aware of AI platforms (as AI platforms are taught by prior art), and the need to filter by licensing is taught (as indicated by teachings for limiting by licensing is taught in relevant claims using prior art), thus an AI platform having associated licensing would mean applying the known task of filtering by license to the AI platform would be obvious as such would have the same motivation to be performed as filtering for licensing for any other thing requiring of licensing [wherein the authorizing comprises creating the subset of the candidate set of training datasets, the subset including those training datasets in the candidate set for which the at least one licensed Al platform identified by the second data block of the respective training dataset includes the identifier of the Al platform]. The application of a known method (filtering in relation to licensing and/or input) to a known element (AI platform) is an obvious step to perform as the motivation to do so would be the same as why one of ordinary skill in the art would apply the filtering in another context.
Claim 24:
The method defined in claim 22 is taught by Weast and Neal.
Weast teaches:
wherein the proposed usage of the Al training data comprises a proposed geographic region, wherein the use rights in the second data block of each of the training datasets further comprises a licensed geographic region, wherein the authorizing comprises creating the subset of the candidate set of training datasets, the subset including those training datasets in the candidate set for which the licensed geographic region in the second data block of the respective training dataset includes the proposed geographic region.
[Weast 0028]: “In example embodiments, the geo-location tag or other location-related metadata may define a geographical area (e.g., that is delimited with a set of points). In example embodiments, the location-related metadata may be identified as corresponding to a region or a country. In example embodiments, the location-related metadata may define a type of the location (e.g., urban versus countryside). In example embodiments, such location-related metadata is added to every piece of training data so that there is traceability from data collection for training through deployment of that model for commercial purposes. In example embodiments, every tag and training data is signed by the entity that generates it. In this way, the entity that generated the data may be validated and it may be validated that the tag has not been tampered with. In example embodiments, the content being generated can be attached to a DRM license managed by each jurisdiction (e.g., including a government or corporation, such as an original design manufacturer (ODM)) to enforce that specific data with specific tags are used in the right place. In example embodiments, depending on a tax agreement with a country, the country may be banned from using certain data or an amount of use of that certain data by the country may be restricted. Alternatively, this could be done by using geographic based certificates [wherein the proposed usage of the Al training data comprises a proposed geographic region, wherein the use rights in the second data block of each of the training datasets further comprises a licensed geographic region, wherein the authorizing comprises creating the subset of the candidate set of training datasets, the subset including those training datasets in the candidate set for which the licensed geographic region in the second data block of the respective training dataset includes the proposed geographic region] that sign the data using a certificate issued by the controlling government.”
Aspects related to created a subset of training data is shown earlier in claims, such as claim 1. Support for the geographic based restrictions for data is found in other elements of the above paragraph, such as “In example embodiments, such location-related metadata is added to every piece of training data so that there is traceability from data collection for training through deployment of that model for commercial purposes.” and “In example embodiments, the content being generated can be attached to a DRM license managed by each jurisdiction (e.g., including a government or corporation, such as an original design manufacturer (ODM)) to enforce that specific data with specific tags are used in the right place.”.
Paragraph 55 of Weast also notes a motivation for limiting by region as additional support for the teachings [Weast 0055]: “In example embodiments, machine learning for fine-grained data tagging is implemented. The training data is tagged with scene description to show multiple features to help the decision on regulated versus un-regulated data. Examples: (1) Data collected on a highway in country1/region1 that is a common highway with country2/region2 can be regulated data that does not need taxation in country2/region2; (2) Data relative to safe driving (e.g., data including traffic signs) may not be subject to taxation; (3) Data pertaining to pedestrians may be regulated or unregulated in some regions based on privacy laws.”
Claim 34:
A non-transitory computer-readable storage medium storing instructions which, when executed by a processing entity of a computing device, causes the computing device to carry out a method build a customized training dataset for training an artificial intelligence (Al) platform, the method comprising: obtaining (i) specifications of training data for training the Al platform and (ii) data indicative of a proposed usage of the training data; consulting a database of training datasets, each associated with use rights, to identify a candidate set of training datasets matching the specifications; and authorizing release of a subset of the training datasets in the candidate set of training datasets based on the data indicative of the proposed usage of the Al training data and the use rights associated with the training data sets in the candidate set of training datasets.
The above limitations are considered taught by teachings used to reject claim 1, as the limitations contain the same subject matter as claim 1, and the rejection for claim 1 includes the teachings of a computer-readable storage medium.
Claim 35:
A computer system for building a customized training dataset for training an artificial intelligence (Al) platform, the computer system comprising: a user interface module configured for obtaining (i) specifications of training data for training the Al platform and (ii) data indicative of a proposed usage of the training data; a filtering module configured for consulting a database of training datasets, each associated with use rights, to identify a candidate set of training datasets matching the specifications; and a rights management module configured for authorizing release of a subset of the training datasets in the candidate set of training datasets based on the data indicative of the proposed usage of the Al training data and the use rights associated with the training data sets in the candidate set of training datasets.
The above limitations are considered taught by teachings used to reject claim 1, as the limitations contain the same subject matter as claim 1.
The modules are seen as taught by the elements of claim 1, as the modules are configured perform the limitations already taught in claim 1.
A computer system is interpreted as having a processor and memory as while claim 35 does not explicitly recite hardware such as memory, processor, or other, the specification paragraph 39 notes figure 27 is a block diagram of a computing system and figure 27 shows memory, processor, and more. Paragraph 2 of the specification indicates “The present disclosure relates to the management of data components used to train Artificial Intelligence (Al) platforms”, which indicates the management server parts of a computing system in paragraph 39 performs the limitations noted in claim 35. A computer system is deemed to be akin to a computing system as paragraph 198 of the specification notes:
[Current Specification 0198] :“It should be appreciated that throughout the specification, discussions utilizing terms such as "processing", "computing", "calculating", "determining", "analyzing" or the like, can refer to the action and/or processes of a computer or computing system, or similar electronic computing device, that manipulate and/or transform data represented as physical, such as electronic, quantities into other data similarly represented as physical quantities.”
Claims 5 and 15 is/are rejected under 35 U.S.C. 103 as being unpatentable over Weast, Neal, and Nowruzi.
Claim 5:
The method defined in claim 1 is taught by Weast and Neal.
Weast does not explicitly teach:
wherein the specifications include a minimum percentage of training data derived from real-world imaging or a maximum percentage of training data derived synthetically
Nowruzi teaches:
wherein the specifications include a minimum percentage of training data derived from real-world imaging or a maximum percentage of training data derived synthetically.
[Nowruzi 4.3 Synthetic-Real Data Mixing page 7]: "We can further see that increasing the real dataset ratio from 2.5% to 5% and 10% [wherein the specifications include a minimum percentage of training data derived from real-world imaging or a maximum percentage of training data derived synthetically] in all cases results in an increase in the performance. These gains could be observed as an increase in precision, recall, or both of the measures."
[Nowruzi 3 Datasets and Data Statistics page 3]: "To perform a comprehensive study on the requirements of having expensive real data, we use a set of real and synthetic datasets."
Aspects of the specification input is taught in claim 1 by Weast and Neal, where Nowruzi is indicating that a percentage of real data is known as a factor for datasets.
One of ordinary skill in the art, prior to the effective filing date, would have been motivated to combine Weast and Nowruzi. Weast and Nowruzi are in the same field of endeavor of information handling such as for training data. One of ordinary skill in the art would have been motivated to combine Weast and Nowruzi to ensure an amount of real data. Nowruzi indicates that the use of percentage ratios of real data be used in training data for performance of the model (as [Nowruzi 4.3 Synthetic-Real Data Mixing page 7] notes the improvement of performance with higher real data percentages, [Nowruzi 2 Literature review page 2] notes there is a point where synthetic data saturates the dataset and does not provide more improvement, and [Nowruzi 2 Literature Review page 2] notes that synthetic data would still want to be used to help overcome the issue with acquiring data).
[Nowruzi 2 Literature review page 2]: "In contrast to (Tremblay et al., 2018), (Abu Alhaija et al., 2018) uses real scenes and augments them with the synthetic objects. This way, the complexity associated with the creation of 3D environment models and spawn probabilities are avoided. It is shown that after augmenting the dataset with certain number of images a saturation point is reached. Cluttering the images with too many objects also reduces the model performance. Another interesting finding of this work is about the significance of proper background and camera noise modeling."
[Nowruzi 2 Literature review page 3]: "The findings from the literature can be concluded in few rules of thumb; more data is almost always better than less data. Samples acquired from the test environment contribute more to the performance of the model. Augmenting real images saturate once a specific ratio is passed. Realistic sensor distortion models and environment distribution models have a larger effect on the final performance than the photo-realism of the samples."
[Nowruzi 2 Literature Review page 2]: "The necessity of large amounts of annotated data is a bottleneck in computer vision tasks. One way of dealing with this issue is to use cheap synthetically generated training images. However, this approach posesses an important issue; how the synthetic and real data should be used to optimize training of a model."
Claim 15:
The method defined in claim 12 is taught by Weast and Neal.
Nowruzi teaches:
wherein the media metadata corresponding to the training media in the first data block is further indicative of whether the training media was derived from real-world imaging or derived synthetically.
[Nowruzi 3 Datasets and Data Statistics page 3]: "To perform a comprehensive study on the requirements of having expensive real data, we use a set of real and synthetic datasets [wherein the media metadata corresponding to the training media in the first data block is further indicative of whether the training media was derived from real-world imaging or derived synthetically as this indicates the data is known to be real or synthetic, such as by being in a dataset indicated to be of real or synthetic data]."
The motivation to use Nowruzi in combination with Weast is the same as in claim 5.
Claims 21 is/are rejected under 35 U.S.C. 103 as being unpatentable over Weast, Neal, and Redlich.
Claim 21:
The method defined in claim 10 is taught by Weast and Neal.
Weast teaches:
wherein the media metadata corresponding to the training media in the first data block comprises information related to a list of objects depicted in the training media in the first data block
[Weast 0055]: “In example embodiments, machine learning for fine-grained data tagging is implemented. The training data is tagged with scene description [wherein the media metadata corresponding to the training media in the first data block comprises information related to a list of objects depicted in the training media in the first data block] to show multiple features to help the decision on regulated versus un-regulated data. Examples: (1) Data collected on a highway in country1/region1 that is a common highway with country2/region2 can be regulated data that does not need taxation in country2/region2; (2) Data relative to safe driving (e.g., data including traffic signs) may not be subject to taxation; (3) Data pertaining to pedestrians may be regulated or unregulated in some regions based on privacy laws.”
Neal teaches:
wherein the specifications comprise a user- specified list of objects
[Neal 0110]: “Referring back to FIG. 11, the right portion side portion of GUI instance 1100 is a data usage selection GUI 1130 (labeled “how are you using the data?”), which is used to specify how the data is being used in the global context 722g and/or local context 722b [wherein the specifications comprise a user- specified list of objects where this is for indicating the user input is taught by Neal, as the relevance of a list of objects is taught more by Weast indicating such is stored for the data and Redlich is used to indicate the searching for such]. The GUI 1130 includes various GCEs 1131 for selecting the data usage(s) for this org context 722. FIG. 14 shows a GUI instance 1400 including a GUI element 1430 for adding and/or removing different data usage taxonomies. In this example, the GUI element 1430 includes prepopulated taxonomies that have a number of different data uses. As can be seen in FIG. 14, all of the taxonomies are hierarchical. The user can also make changes to how the taxonomies are listed and arranged within the hierarchy to fit their particular org 120.”
The motivation to combine with Neal is the same as claim 1.
Weast does not explicitly teach:
wherein to identify a candidate set of training datasets matching the specifications, the method further comprises comparing the user-specified list of objects to the list of objects in the third data block of each of the training datasets, the candidate set of training datasets excluding those training datasets for which the list of objects in the third data block of the respective training dataset does not contain any of the objects in the user-specified list of objects
Redlich teaches:
wherein to identify a candidate set of training datasets matching the specifications, the method further comprises comparing the user-specified list of objects to the list of objects in the third data block of each of the training datasets, the candidate set of training datasets excluding those training datasets for which the list of objects in the third data block of the respective training dataset does not contain any of the objects in the user-specified list of objects
[Redlich Column 12 Line 29]: "Data 1B is also applied to control 52. The SC filter 45 generates a result command 51 or an output control signal to indicate the presence or absence of certain SC data in data block 1B [wherein to identify a candidate set of training datasets matching the specifications, the method further comprises comparing the user-specified list of objects to the list of objects in the third data block of each of the training datasets, the candidate set of training datasets excluding those training datasets for which the list of objects in the third data block of the respective training dataset does not contain any of the objects in the user-specified list of objects as this indicates the use of content and filtering for content]. Contextual relevance or taxonomic key words or terms may also trigger filter result output 51. This signal is applied to control 52 which activates data process 54. Some data processes are copy, extract, distribute, archive, or destroy. Further, as discussed below, SC filter 45 may be set as a screening filter which is then used to identify a certain categorical filter which then triggers the data process. Alternatively, the data input 1A or 1B may come with a data process request and the data must be filtered or scrubbed before the data process is carried out by the enterprise computer system."
One of ordinary skill in the art, prior to the effective filing date, would have been motivated to combine Weast and Redlich. Weast and Redlich are in the same field of endeavor of information management or filtering. One of ordinary skill in the art would have been motivated to combine Weast and Redlich to filter by content in order to perform processes or such with the wanted content ([part of Redlich Column 12 Line 29]: “Contextual relevance or taxonomic key words or terms may also trigger filter result output 51. This signal is applied to control 52 which activates data process 54. Some data processes are copy, extract, distribute, archive, or destroy. Further, as discussed below, SC filter 45 may be set as a screening filter which is then used to identify a certain categorical filter which then triggers the data process. Alternatively, the data input 1A or 1B may come with a data process request and the data must be filtered or scrubbed before the data process is carried out by the enterprise computer system."). Weast and other references like Neal indicate the premise of filtering (as indicated by the limiting by scope and such as shown in other claims), but to provide indication of a more general searching for content the reference Redlich is utilized as Redlich indicates the premise of filtering for content to use. The combination of Redlich with Weast would allow the storing of object related data from Weast in data collection/harvesting to be used to filter to that content by searching or filtering to said content.
Below shows more detail about filtering to select content such as how it can involve contextual filters or is representable by things like words, elements or objects (which can be used to indicate objects of selection being used for searching)
[Redlich Column 12 Line 43]: "Some key operations of the system and process follow. The invention organizes and processes data important to an enterprise operating the distributed computing system. The SC or select content is represented by one or more predetermined words, characters, images, data elements or data objects. The computing system has a plurality of select content data stores 7a-7e for respective ones of a plurality of enterprise designated categorical filters which include content-based filters, contextual filters and taxonomic classification filters, all operatively coupled over a communications network. An archive filter 8a operates with data at rest 2c at the top left of lifecycle 1. Aged content is subject to archiving policies or back-up policies. A data input 1A or 1B is processed through at least one activated categorical SC filter 45 to obtain select content, and contextually associated select content and taxonomically associated select content (SC-la) as aggregated select content. The aggregated select content is stored in the corresponding select content data store 47 and, more specifically, filter aggregate file 48. A data process 54 from the group of data processes including a copy process, a data extract process, a data archive process, a data distribution process and a data destruction process is associated with the activated categorical SC filter 45 and the method and system applies the associated data process 54 to a further data input 1B based upon a result 51 of that further data 1B being processed by the activated categorical SC filter 45 utilizing the aggregated select content data per path 49."
Below quote shows relevance with being a management system for how information gets used and then another quote indicating it is related to things such as searching in the second quote:
[Redlich Column 8 Line 36]: "Information Rights Management (IR Mgt) can and is presently imposed on data files with the use of data file metadata, tags or labels. For example, the file property may be set at “Read Only” which limits revision rights. However, the content in information files is not subject to many controls. IR Mgt (a) controls the open, print and edit functions; (b) limits access, permissions, and distribution; (c) establishes and permits changes to permission levels; and, (d) provides and audit trail. Since the value of the information changes over the life cycle of the information file, oftentimes due to events and objects outside of the information file itself, it is the content that should be managed and subjected to an information infrastructure."
[Redlich Column 17 Line 10]: "Different tools enable selection of granular data items for different activities which may include, among other things, their use as: (1) key words to search engines; (2) headings for granular data elements, meta tags for such elements; (3) a base for build up of expansive content (supple docs); and (4) filling in missing information from documents (data mining third party sources."
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
Kumar et al (US20200342014) is relevant art that teaches data indexing for tagging and such to help find datasets. Filtering or searching methods for training data is similar to the aspects in the current invention.
Lally et al (US 20220414237 A1) is relevant art that that seeks to make training data or information more accessible, searchable, and secure to help better the ability to train AI models. The motivations in Lally appear similar to that of the current invention.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to CHRISTOPHER D DEVORE whose telephone number is (703)756-1234. The examiner can normally be reached Monday-Friday 7:30 am - 5 pm EST.
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, Michael J Huntley can be reached at (303) 297-4307. 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.
/C.D.D./Examiner, Art Unit 2129
/MICHAEL J HUNTLEY/Supervisory Patent Examiner, Art Unit 2129
Prosecution Insights