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 .
Continued Examination Under 37 CFR 1.114
A request for continued examination under 37 CFR 1.114, including the fee set forth in 37 CFR 1.17(e), was filed in this application after final rejection. Since this application is eligible for continued examination under 37 CFR 1.114, and the fee set forth in 37 CFR 1.17(e) has been timely paid, the finality of the previous Office action has been withdrawn pursuant to 37 CFR 1.114. Applicant's submission filed on 1/13/2025 has been entered.
Amendments
Claims 1, 7, 12, 18, and 20 are amended. Claim 3 is cancelled. Claims 1, 4-13, 18, and 20 are pending and have been considered.
Claim Rejections - 35 USC § 112
The following is a quotation of 35 U.S.C. 112(b):
(b) CONCLUSION.—The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the inventor or a joint inventor regards as the invention.
The following is a quotation of 35 U.S.C. 112 (pre-AIA ), second paragraph:
The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the applicant regards as his invention.
Claims 9-11 are 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.
Claim 9 recites the limitation "the stop condition". There is insufficient antecedent basis for this limitation in the claim. For examining purposes, Examiner is interpreting the claim as if it had recited “a stop condition”.
Claim 10, which depends on claim 9, fails to cure the issue in claim 9 and is therefore
rejected under the same rationale.
Claim 11 recites the limitation "the stop condition". There is insufficient antecedent basis for this limitation in the claim. For examining purposes, Examiner is interpreting the claim as if it had recited “a stop condition”.
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, 4-13, 18, and 20 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more.
Step 1: Claims 1 and 4-13 are method claims for growing labels from semi-supervised learning. Claims 18 and 20 are computer program product claims for growing labels from semi-supervised learning. Therefore, claims 1, 4-13, 18, and 20 are directed to either a process, machine, manufacture, or composition of matter.
Claim 1:
Step 2A Prong 1: The claim recites the following limitations:
associating a first probability distribution to each labeled data item in the collection of labeled data, the first probability distribution including one probability value for each label in the set of labels (Mental process of evaluation and judgement which can be reasonably performed in one’s mind or with the aid of pencil and paper – a person could manually evaluate a collection of labeled data and a set of labels, and then associate a probability distribution to each labeled data item in the collection including one probability value for each label in the set of labels)
associating a second probability distribution to each unlabeled data item in the collection of unlabeled data (Mental process of evaluation and judgement which can be reasonably performed in one’s mind or with the aid of pencil and paper – a person could manually evaluate unlabeled data items in a collection of unlabeled data and then associate a probability distribution to each unlabeled data item)
labelling the unlabeled data Mental process of evaluation and judgement which can be reasonably performed in one’s mind or with the aid of pencil and paper – a person could manually evaluate unlabeled data and then label the unlabeled data by indicating a probability of a particular known label being associated with a particular unlabeled data)
associating a label to each unlabeled data item utilizing a peaking probability distribution (Mental process of evaluation and judgement which can be reasonably performed in one’s mind or with the aid of pencil and paper – a person could manually evaluate unlabeled data items and a peaking probability distribution, and the associate a label to each unlabeled data item utilizing a peaking probability distribution)
Accordingly, the claim recites an abstract idea.
Step 2A Prong 2: The judicial exceptions are not integrated into a practical application. The claim recites
the following additional elements:
receiving … a collection of unlabeled data, the unlabeled data not labelled by manual labelling (Adding insignificant extra-solution activity of mere data gathering to the judicial exception – see MPEP 2106.05(g))
receiving … a collection of labeled data, each labeled data item in the collection being associated with a label in a set of labels by manual labelling … (Adding insignificant extra-solution activity of mere data gathering to the judicial exception – see MPEP 2106.05(g))
… the collection of labelled data smaller than the collection of unlabeled data (Amounts to generally linking the abstract ideas to a particular technological environment or field of use, as discussed in MPEP 2106.05(h))
training the classifier of a machine learning system using the collection of labeled data and labelled previously unlabeled data items, the classifier resulting in having improved classification accuracy versus classification accuracy by training the classifier with only the collection of labeled data labeled by manual labelling (Merely training, or applying a machine learning model on labeled training samples represents adding the words “apply it” (or an equivalent) with the judicial exception, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea -see MPEP 2106.05(f))
… by the autoencoder architecture including one or more autoencoders by tuning parameters in the one or more autoencoders, the one or more autoencoders … (Adding the words “apply it” (or an equivalent) with the judicial exception, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea, as discussed in MPEP 2106.05(f))
… by an/the encoder of an/the autoencoder architecture … (Adding the words “apply it” (or an equivalent) with the judicial exception, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea, as discussed in MPEP 2106.05(f))
Accordingly, the claim is directed to an abstract idea.
Step 2B: The claim does not include additional elements that are sufficient to amount to significantly
more than the judicial exception. The claim recites the following additional elements:
receiving … a collection of unlabeled data, the unlabeled data not labelled by manual labelling (MPEP 2106.05(d)(II) indicates that merely “receiving or transmitting data over a network” is a well-understood, routine, conventional function when it is claimed in a merely generic manner (as it is in the present claim))
receiving … a collection of labeled data, each labeled data item in the collection being associated with a label in a set of labels by manual labelling … (MPEP 2106.05(d)(II) indicates that merely “receiving or transmitting data over a network” is a well-understood, routine, conventional function when it is claimed in a merely generic manner (as it is in the present claim))
… the collection of labelled data smaller than the collection of unlabeled data (Amounts to generally linking the abstract ideas to a particular technological environment or field of use, as discussed in MPEP 2106.05(h))
training the classifier of a machine learning system using the collection of labeled data and labelled previously unlabeled data items, the classifier resulting in having improved classification accuracy versus classification accuracy by training the classifier with only the collection of labeled data labeled by manual labelling (Merely training, or applying a machine learning model on labeled training samples represents adding the words “apply it” (or an equivalent) with the judicial exception, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea -see MPEP 2106.05(f))
… by the autoencoder architecture including one or more autoencoders by tuning parameters in the one or more autoencoders, the one or more autoencoders … (Adding the words “apply it” (or an equivalent) with the judicial exception, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea, as discussed in MPEP 2106.05(f))
… by an/the encoder of an/the autoencoder architecture … (Adding the words “apply it” (or an equivalent) with the judicial exception, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea, as discussed in MPEP 2106.05(f))
The claim is not patent eligible.
Claim 4 incorporates the rejection of claim 1.
Step 2A Prong 1: The judicial exceptions of claim 1 are incorporated. The claim recites: encoding and compressing a particular data item received Mental process of evaluation and judgement which can be reasonably performed in one’s mind or with the aid of pencil and paper – a person could manually evaluate a particular data item and then encode and compress the data item to a compressed data code version of the particular data item); decoding and expanding the compressed data code version to a reconstructed version of the particular data item which is provided at Mental process of evaluation and judgement which can be reasonably performed in one’s mind or with the aid of pencil and paper – a person could manually evaluate a compressed data code version and then decode and evaluate the compressed data code version to a reconstructed version of the particular data item); comparing the output reconstructed version to the input particular data item (Mental process of evaluation and judgement which can be reasonably performed in one’s mind or with the aid of pencil and paper – a person could manually evaluate a reconstructed version and an input particular data item, and then make a comparison between the reconstructed version and the input particular data item); providing, based on the comparison, a loss of information value representing a loss of information from processing the input particular data item to the output reconstructed version, … (Mental process of evaluation and judgement which can be reasonably performed in one’s mind or with the aid of pencil and paper – a person could manually evaluate a comparison and then provide a loss of information value based on the comparison representing a loss of information from processing the input particular data item to the output reconstructed version). Accordingly, the claim recites an abstract idea.
Step 2A Prong 2: The judicial exceptions are not integrated into a practical application. The claim recites
the following additional elements: at an input/output of each autoencoder (Adding the words “apply it” (or an equivalent) with the judicial exception, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea, as discussed in MPEP 2106.05(f)); … where the each autoencoder processes most accurately, with lowest loss of information, a particular data item that is likely a member of one of the one or more classified labeled sets of data that is associated with the each autoencoder and which is associated with one label in the set of labels (Amounts to generally linking the abstract ideas to a particular technological environment or field of use, as discussed in MPEP 2106.05(h)). Accordingly, the additional elements alone or in combination do not integrate the abstract ideas into a practical application because they do not impose any meaningful limits on practicing the abstract ideas. The claim is directed to an abstract idea.
Step 2B: The claim does not include additional elements that are sufficient to amount to significantly
more than the judicial exceptions. At an input/output of each autoencoder (Adding the words “apply it” (or an equivalent) with the judicial exception, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea, as discussed in MPEP 2106.05(f)); where the each autoencoder processes most accurately, with lowest loss of information, a particular data item that is likely a member of one of the one or more classified labeled sets of data that is associated with the each autoencoder and which is associated with one label in the set of labels (Amounts to generally linking the abstract ideas to a particular technological environment or field of use, as discussed in MPEP 2106.05(h)). Accordingly, the additional elements alone or in combination do not integrate the abstract ideas into a practical application because they do not impose any meaningful limits on practicing the abstract ideas. The claim is directed to an abstract idea. The claim is not patent eligible.
Claim 5 incorporates the rejection of claim 1.
Step 2A Prong 1: The judicial exceptions of claim 1 are incorporated. The claim recites: determining, Mental process of evaluation and judgement which can be reasonably performed in one’s mind or with the aid of pencil and paper – a person could manually evaluate a peaking probability distribution and a threshold value, and then determine whether the highest probability in the distribution is above the threshold value, and if so, add to a set of classified labeled data associated with the label the processed unlabeled data item that has the label associated therewith). Accordingly, the claim recites an abstract idea.
Step 2A Prong 2: The judicial exceptions are not integrated into a practical application. The claim recites
the following additional elements: with the computer processing system (Adding the words “apply it” (or an equivalent) with the judicial exception, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea, as discussed in MPEP 2106.05(f)). Accordingly, the additional elements alone or in combination do not integrate the abstract ideas into a practical application because they do not impose any meaningful limits on practicing the abstract ideas. The claim is directed to an abstract idea.
Step 2B: The claim does not include additional elements that are sufficient to amount to significantly
more than the judicial exceptions. With the computer processing system (Adding the words “apply it” (or an equivalent) with the judicial exception, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea, as discussed in MPEP 2106.05(f)). The claim is not patent eligible.
Claim 6 incorporates the rejection of claim 5.
Step 2A Prong 1: The judicial exceptions of claim 5 are incorporated. The claim recites an abstract idea.
Step 2A Prong 2: The judicial exceptions are not integrated into a practical application. The claim recites
the following additional elements: wherein the high probability threshold value is at least 75% probability (.75) (Amounts to generally linking the abstract ideas to a particular technological environment or field of use, as discussed in MPEP 2106.05(h)). Accordingly, the additional elements alone or in combination do not integrate the abstract ideas into a practical application because they do not impose any meaningful limits on practicing the abstract ideas. The claim is directed to an abstract idea.
Step 2B: The claim does not include additional elements that are sufficient to amount to significantly
more than the judicial exceptions. Wherein the high probability threshold value is at least 75% probability (.75) (Amounts to generally linking the abstract ideas to a particular technological environment or field of use, as discussed in MPEP 2106.05(h)). The claim is not patent eligible.
Claim 7 incorporates the rejection of claim 1.
Step 2A Prong 1: The judicial exceptions of claim 1 are incorporated. The claim recites: monitoring, Mental process of evaluation and judgement which can be reasonably performed in one’s mind or with the aid of pencil and paper – a person could manually evaluate a history of label purity values associated with each processed unlabeled data item and determine whether the purity values are not increasing over one or more iterations of processing the unlabeled data items). Accordingly, the claim recites an abstract idea.
Step 2A Prong 2: The judicial exceptions are not integrated into a practical application. The claim recites
the following additional elements: … with the autoencoder architecture … (Adding the words “apply it” (or an equivalent) with the judicial exception, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea, as discussed in MPEP 2106.05(f)). Accordingly, the additional elements alone or in combination do not integrate the abstract ideas into a practical application because they do not impose any meaningful limits on practicing the abstract ideas. The claim is directed to an abstract idea.
Step 2B: The claim does not include additional elements that are sufficient to amount to significantly
more than the judicial exceptions. … with the autoencoder architecture … (Adding the words “apply it” (or an equivalent) with the judicial exception, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea, as discussed in MPEP 2106.05(f)). The claim is not patent eligible.
Claim 8 incorporates the rejection of claim 7.
Step 2A Prong 1: The judicial exceptions of claim 7 are incorporated. The claim recites: monitoring, Mental process of evaluation and judgement which can be reasonably performed in one’s mind or with the aid of pencil and paper – a person could manually evaluate a history of label purity values associated with each processed unlabeled data item and a threshold value, and then determine whether the purity values are not increasing over a threshold number of iterations of processing the unlabeled data items). Accordingly, the claim recites an abstract idea.
Step 2A Prong 2: The judicial exceptions are not integrated into a practical application. The claim recites
the following additional elements: with the autoencoder architecture (Adding the words “apply it” (or an equivalent) with the judicial exception, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea, as discussed in MPEP 2106.05(f)). Accordingly, the additional elements alone or in combination do not integrate the abstract ideas into a practical application because they do not impose any meaningful limits on practicing the abstract ideas. The claim is directed to an abstract idea.
Step 2B: The claim does not include additional elements that are sufficient to amount to significantly
more than the judicial exceptions. With the autoencoder architecture (Adding the words “apply it” (or an equivalent) with the judicial exception, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea, as discussed in MPEP 2106.05(f)). The claim is not patent eligible.
Claim 9 incorporates the rejection of claim 1.
Step 2A Prong 1: The judicial exceptions of claim 1 are incorporated. The claim recites: monitoring, Mental process of evaluation and judgement which can be reasonably performed in one’s mind or with the aid of pencil and paper – a person could manually evaluate a history of label purity values associated with each processed unlabeled data item and determine whether the purity values decrease over one or more iterations of processing the unlabeled data items). Accordingly, the claim recites an abstract idea.
Step 2A Prong 2: The judicial exceptions are not integrated into a practical application. The claim recites
the following additional elements: with the autoencoder architecture (Adding the words “apply it” (or an equivalent) with the judicial exception, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea, as discussed in MPEP 2106.05(f)). Accordingly, the additional elements alone or in combination do not integrate the abstract ideas into a practical application because they do not impose any meaningful limits on practicing the abstract ideas. The claim is directed to an abstract idea.
Step 2B: The claim does not include additional elements that are sufficient to amount to significantly
more than the judicial exceptions. With the autoencoder architecture (Adding the words “apply it” (or an equivalent) with the judicial exception, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea, as discussed in MPEP 2106.05(f)). The claim is not patent eligible.
Claim 10 incorporates the rejection of claim 9.
Step 2A Prong 1: The judicial exceptions of claim 9 are incorporated. The claim recites: monitoring, Mental process of evaluation and judgement which can be reasonably performed in one’s mind or with the aid of pencil and paper – a person could manually evaluate a history of label purity values associated with each processed unlabeled data item and a threshold number, and then determine whether the purity values decrease over a threshold number of iterations of processing the unlabeled data items). Accordingly, the claim recites an abstract idea.
Step 2A Prong 2: The judicial exceptions are not integrated into a practical application. The claim recites
the following additional elements: with the autoencoder architecture (Adding the words “apply it” (or an equivalent) with the judicial exception, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea, as discussed in MPEP 2106.05(f)). Accordingly, the additional elements alone or in combination do not integrate the abstract ideas into a practical application because they do not impose any meaningful limits on practicing the abstract ideas. The claim is directed to an abstract idea.
Step 2B: The claim does not include additional elements that are sufficient to amount to significantly
more than the judicial exceptions. With the autoencoder architecture (Adding the words “apply it” (or an equivalent) with the judicial exception, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea, as discussed in MPEP 2106.05(f)). The claim is not patent eligible.
Claim 11 incorporates the rejection of claim 1.
Step 2A Prong 1: The judicial exceptions of claim 1 are incorporated. The claim recites: monitoring, Mental process of evaluation and judgement which can be reasonably performed in one’s mind or with the aid of pencil and paper – a person could manually evaluate a history of label purity values associated with each processed unlabeled data item and determine whether the purity values are not increasing over one or more iterations of processing the unlabeled data items). Accordingly, the claim recites an abstract idea.
Step 2A Prong 2: The judicial exceptions are not integrated into a practical application. The claim recites
the following additional elements: with the autoencoder architecture (Adding the words “apply it” (or an equivalent) with the judicial exception, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea, as discussed in MPEP 2106.05(f)). Accordingly, the additional elements alone or in combination do not integrate the abstract ideas into a practical application because they do not impose any meaningful limits on practicing the abstract ideas. The claim is directed to an abstract idea.
Step 2B: The claim does not include additional elements that are sufficient to amount to significantly
more than the judicial exceptions. With the autoencoder architecture (Adding the words “apply it” (or an equivalent) with the judicial exception, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea, as discussed in MPEP 2106.05(f)). The claim is not patent eligible.
Claim 12 incorporates the rejection of claim 1.
Step 2A Prong 1: The judicial exceptions of claim 1 are incorporated. The claim recites: in response to the autoencoder architecture detecting a stop condition, Mental process of evaluation and judgement which can be reasonably performed in one’s mind or with the aid of pencil and paper – a person could manually evaluate a stop condition, a peaking probability distribution, and a threshold value, and in response to detecting a stop condition, associate a label to an unlabeled data item based on the label being associated with a highest probability value in the peaking probability distribution and the highest probability exceeding the threshold value). Accordingly, the claim recites an abstract idea.
Step 2A Prong 2: The judicial exceptions are not integrated into a practical application. The claim recites
the following additional elements: the autoencoder architecture automatically (Adding the words “apply it” (or an equivalent) with the judicial exception, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea, as discussed in MPEP 2106.05(f)). Accordingly, the additional elements alone or in combination do not integrate the abstract ideas into a practical application because they do not impose any meaningful limits on practicing the abstract ideas. The claim is directed to an abstract idea.
Step 2B: The claim does not include additional elements that are sufficient to amount to significantly
more than the judicial exceptions. The autoencoder architecture automatically (Adding the words “apply it” (or an equivalent) with the judicial exception, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea, as discussed in MPEP 2106.05(f)). The claim is not patent eligible.
Claim 13 incorporates the rejection of claim 12.
Step 2A Prong 1: The judicial exceptions of claim 12 are incorporated. The claim recites an abstract idea.
Step 2A Prong 2: The judicial exceptions are not integrated into a practical application. The claim recites
the following additional elements: wherein the high probability threshold value is at least 90% probability (.9) (Amounts to generally linking the abstract ideas to a particular technological environment or field of use, as discussed in MPEP 2106.05(h)). Accordingly, the additional elements alone or in combination do not integrate the abstract ideas into a practical application because they do not impose any meaningful limits on practicing the abstract ideas. The claim is directed to an abstract idea.
Step 2B: The claim does not include additional elements that are sufficient to amount to significantly
more than the judicial exceptions. Wherein the high probability threshold value is at least 90% probability (.9) (Amounts to generally linking the abstract ideas to a particular technological environment or field of use, as discussed in MPEP 2106.05(h)). The claim is not patent eligible.
Claim 18
Step 2A Prong 1: Claim 18 comprises limitations similar to those of claim 1. The claim recites an abstract idea.
Step 2A Prong 2: The judicial exceptions are not integrated into a practical application. The claim recites the following additional elements: a non-transitory computer readable storage medium readable by a processing device and storing program instructions for execution by the processing device, said program instructions comprising […]. These additional elements are all computer components. They are recited at a high level of generality such that they amount to no more than mere instructions to apply the exception using generic computer components, as discussed in MPEP2106.05(f). Accordingly, the additional elements do not integrate the abstract ideas into a practical application because they do not impose any meaningful limits on practicing the abstract ideas. The claim is directed to an abstract idea.
Step 2B: The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exceptions. A non-transitory computer readable storage medium readable by a processing device and storing program instructions for execution by the processing device, said program instructions comprising … amounts to no more than mere instructions to apply the exception using generic computer components, as discussed in MPEP 2106.05(f). The claim is not patent eligible.
Claim 20 comprise limitations similar to those of claim 12 and is therefore rejected for at least the same rationale.
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.
9. Claims 1, 4, 7, 11, and 18 are rejected under 35 U.S.C. 103 as being unpatentable over Li et al. “Two-View Label Propagation to Semi-supervised Reader Emotion Classification” 2016 (hereinafter Li) in view of Choi et al. Pub. No. US 20200372368 A1 (hereinafter Choi).
Regarding claim 1, Li teaches: A computer-implemented method for automatically labeling an amount of unlabeled data for training one or more classifiers of a machine learning system to improve classification accuracy of the one or more classifiers with a larger amount of accurately labeled data and avoid the need for manual classification of training data, the method comprising: receiving a collection of unlabeled data, the unlabeled data not labelled by manual labelling (Page 2649, Section 3, Paragraph 1 and Page 2650, Figure 4, 1a teach that a collection of unlabeled data (i.e., not labelled by manual labelling) has been received);
receiving a collection of labeled data, each labeled data item in the collection being associated with a label in a set of labels by manual labelling, the collection of labelled data smaller than the collection of unlabeled data (Page 2649, Section 3, Paragraph 1 and Page 2650, Figure 4, 1b teach that a collection of labeled data has been received and that the labeled data is associated with one of the two emotion categories (i.e., either positive or negative emotion); Page 2652, Section 5.1, Data collection sub-section teaches that the labels are derived from voted emotion tags (i.e., manual labelling) and the Data setting sub-section teaches that 0.5%, 1%, and 2% of the data is selected as initial labeled data and the remaining data as unlabeled data (i.e., the collection of labelled data is smaller than the collection of unlabeled data));
associating a first probability distribution to each labeled data item in the collection of labeled data, the first probability distribution including one probability value for each label in the set of labels (Page 2650, Figure 4, 1a teaches assigning each labeled sample with a probability distribution (1,0) or (0,1) according to its label r);
associating a second probability distribution to each unlabeled data item in the collection of unlabeled data (Page 2650, Figure 4, 1b teaches assigning each unlabeled sample with an initial probability distribution (0.5, 0.5));
labelling the unlabeled data indicating a probability of a particular known label being associated with a particular unlabeled data (Page 2650, Figure 4.2 and Page 2651, Paragraph 2 teach that an initial probability is assigned to each unlabeled sample for each possible label (i.e., a probability of a given label being associated with that unlabeled sample), and the procedure iterates in Step 2 until the probability matrix converges);
associating a label to each unlabeled data item utilizing a peaking probability distribution (Figure 4, Step 2 teaches that the procedure loops until the probability matrix converges and Figure 4, Step 3 teaches that each unlabeled instance is then assigned a label by computing argmax pir (i.e., associating a label to each unlabeled document by utilizing a peaking probability distribution));
training the classifier of a machine learning system using the collection of labeled data and labelled previously unlabeled data items, the classifier resulting in having improved classification accuracy versus classification accuracy by training the classifier with only the collection of labeled data labeled by manual labelling (Page 2653, Section 5.2 and Figure 7 teach applying label propagation (i.e., label previously unlabeled data items) to perform semi-supervised learning (i.e., training a classifier) to improve classifier accuracy).
Li does not explicitly teach: receiving by an encoder of an autoencoder architecture a collection of unlabeled data; receiving by the encoder of the autoencoder architecture a collection of labeled data; labelling the unlabeled data by the autoencoder architecture including one or more autoencoders by tuning parameters in the one or more autoencoders, the one or more autoencoders indicating a probability […].
However, in the analogous art, Choi teaches: receiving by an encoder of an autoencoder architecture a collection of unlabeled data ([0025] teaches that the learning data used to train the semi-supervised learning apparatus that includes a plurality of autoencoders is comprised of both unlabeled and labeled data; Figures 1 and 2 teach that the encoders of the autoencoders receive the extracted feature values of the input data);
receiving by the encoder of the autoencoder architecture a collection of labeled data ([0025] teaches that the learning data used to train the semi-supervised learning apparatus that includes a plurality of autoencoders is comprised of both unlabeled and labeled data; Figures 1 and 2 teach that the encoders of the autoencoders receive the extracted feature values of the input data);
labelling the unlabeled data by the autoencoder architecture including one or more autoencoders by tuning parameters in the one or more autoencoders, the one or more autoencoders indicating a probability of a particular known label being associated with a particular unlabeled data ([0041-0042] teaches using a plurality of autoencoders to learn unlabeled data and in doing so, only a result of one encoder of the plurality of encoders approaches zero; [0064] teaches that results of the encoders may be compared and the target class of the autoencoder whose result is closest to zero may be output as a final label result for a piece of data (i.e., the one or more autoencoders indicate a probability of a particular known label being associated with a particular unlabeled data, as [0051] teaches that the probability of the corresponding class increases as the result of the encoder is close to zero); [0011-0013] teaches that the encoders may be learned to generate desired encoding outputs (i.e., by tuning parameters in the one or more autoencoders)).
It would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to have modified the invention of Li with the above teaching of Choi because doing so would lead to an expected improvement in classification performance when performing supervised learning utilizing pieces of unlabeled data which are remaining even when only some pieces of data of an entire dataset are labeled (Choi, [0004] and [0090]).
Regarding claim 4, the combination of Li and Choi teaches all of the elements of claim 1 as shown in the rejection above. Choi further teaches: wherein the processing, with the autoencoder architecture, each unlabeled data item, comprises: encoding and compressing a particular data item received at an input of each autoencoder to a compressed data code version of the particular data item ([0047] and Figure 2 teach that the encoders compress the input information into an encoding value (i.e., a compressed data code version));
decoding and expanding the compressed data code version to a reconstructed version of the particular data item which is provided at an output of the each autoencoder ([0030], [0047], and Figure 2 teach that the decoder receives the encoding value from the encoder (i.e., the compressed data code version of the input information) and outputs the same value as the input information (i.e., a reconstructed version of the input information), which implies that the decoder decodes and expands the compressed version of the input information received from the encoder);
comparing the output reconstructed version to the input particular data item ([0060] teaches comparing the output value of the decoder (i.e., the output reconstructed version) to the input value of the encoder (i.e., the input particular data item)); and
providing, based on the comparison, a loss of information value representing a loss of information from processing the input particular data item to the output reconstructed version ([0060] teaches providing a calculated output value L2 based on the comparison of decoder output to encoder input (i.e., a loss of information from processing the input data item to the output reconstructed version),
where the each autoencoder processes most accurately, with lowest loss of information, a particular data item that is likely a member of one of the one or more classified labeled sets of data that is associated with the each autoencoder and which is associated with one label in the set of labels ([0042] teaches that each of the autoencoders performs learning on each encoder so that only a result of one encoder of the plurality of encoders approaches zero and results of the remaining encoders are farther from zero due to marginal entropy loss of the encoding values output from the plurality of encoders being minimized using the marginal entropy loss).
Regarding claim 7, the combination of Li and Choi teaches all of the elements of claim 1 as shown in the rejection above. Choi further teaches: further comprising a stop condition associated with the autoencoder architecture wherein the stop condition comprises: monitoring, with the autoencoder architecture, a history of label probability purity values associated with the processed each unlabeled data item not increasing over one or more iterations of processing unlabeled data items by the autoencoder architecture ([0042] teaches that only a result of one encoder of the plurality of encoders approaches zero, or converges to zero as taught by [0053], (i.e., a stop condition) and results of the remaining encoders are farther from zero; [0046] teaches calculating a label probability that is inversely based on this encoder result value; this implies that the calculated probability of the single encoder whose result converges to zero (i.e., a label probability purity value) does not increase over the previous iterations of processing (i.e., a history of label probability purity values not increasing over one or more iterations)).
Regarding claim 11, the combination of Li and Choi teaches all of the elements of claim 1 as shown in the rejection above. Choi further teaches: wherein the stop condition comprises: monitoring, with the autoencoder architecture, a history of label probability purity values associated with the processed each unlabeled data item not increasing over one or more iterations of processing unlabeled data items by the autoencoder architecture ([0042] teaches that only a result of one encoder of the plurality of encoders approaches zero, or converges to zero as taught by [0053], (i.e., a stop condition) and results of the remaining encoders are farther from zero; [0046] teaches calculating a label probability that is inversely based on this encoder result value; this implies that the calculated probability of the single encoder whose result converges to zero (i.e., a label probability purity value) does not increase over the previous iteration of processing (i.e., a history of label probability purity values not increasing over one or more iterations)).
Regarding claim 18, it is a computer program product claim comprising limitations similar to those of claim 1 and is therefore rejected for at least the same rationale.
10. Claims 5-6, 12-13, and 20 are rejected under 35 U.S.C. 103 as being unpatentable over Li in view of Choi, as applied to claims 1 and 18 above, and in further view of Pedro et al. Pub. No. US 20090024615 A1 (hereinafter Pedro).
Regarding claim 5, the combination of Li and Choi teaches all of the elements of claim 1 as shown in the rejection above, but does not explicitly teach: determining, with the computer processing system, whether a highest probability in a peaking probability distribution associated with one processed unlabeled data item is above a high probability threshold value, and in response automatically adding to the set of classified labeled data associated with the label a new labeled data item which is the processed unlabeled data item that has the label automatically associated therewith.
However, in the analogous art, Pedro teaches: determining, with the computer processing system, whether a highest probability in a peaking probability distribution associated with one processed unlabeled data item is above a high probability threshold value, and in response automatically adding to the set of classified labeled data associated with the label a new labeled data item which is the processed unlabeled data item that has the label automatically associated therewith ([0014] teaches adjusting the label probabilities of an edge and then assigning a label with a maximum probability to the edge if said maximum probability is greater than a predetermined threshold to create a labeled graph (i.e., representing the unlabeled data items that have now been automatically associated with the label and added to the set of classified labeled data)).
It would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to have modified the combination of Li and Choi with the above teaching of Pedro because doing so would lead to an expected increase in the ability to achieve promising performance in terms of precision and recall results (Pedro, [0100-0101]).
Regarding claim 6, the combination of Li, Choi, and Pedro teaches all of the elements of claim 5 as shown in the rejection above. Pedro further teaches: wherein the high probability threshold value is at least 75% probability (.75) ([0100] teaches that the threshold for label assignment may be .9 (i.e., at least 75% probability)).
Regarding claim 12, the combination of Li and Choi teaches all of the elements of claim 1 as shown in the rejection above. Choi further teaches: in response to the autoencoder architecture detecting a stop condition, the autoencoder architecture automatically associating a label in the set of labels to the processed unlabeled data item, based on the label being associated with a highest probability value in a peaking probability distribution associated with the processed unlabeled data item and ([0041-0042] teaches that only a result of one encoder of the plurality of encoders approaches zero (i.e., a stop condition); [0064] teaches that results of the encoders may be compared and the target class of the autoencoder whose result is closest to zero may be output as a final result (i.e., automatically associating a label in the set of labels to the processed unlabeled data item based on the label being associated with a highest probability value in a peaking probability distribution, as [0051] teaches that the probability of the corresponding class increases as the result of the encoder is close to zero)).
The combination of Li and Choi does not explicitly teach: … the highest probability exceeding a high probability threshold value.
However, in the analogous art, Pedro teaches: … associating a label in the set of labels to the processed unlabeled data item, based on the label being associated with a highest probability value … and the highest probability exceeding a high probability threshold value ([0014] teaches adjusting the label probabilities of an edge and then assigning a label with a maximum probability to the edge if said maximum probability is greater than a predetermined threshold to create a labeled graph).
It would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to have modified the combination of Li and Choi with the above teaching of Pedro because doing so would lead to an expected increase in the ability to achieve promising performance in terms of precision and recall results (Pedro, [0100-0101]).
Regarding claim 13, the combination of Li, Choi, and Pedro teaches all of the elements of claim 12 as shown in the rejection above. Pedro further teaches: wherein the high probability threshold value is at least 90% probability (.9) ([0100] teaches using a high probability threshold value of .9 to determine whether to assign a label to an unlabeled data item).
Regarding claim 20, it is a computer program product claim comprising limitations similar to those of claim 12 and is therefore rejected for at least the same rationale.
11. Claims 8-10 are rejected under 35 U.S.C. 103 as being unpatentable over Li in view of Choi, as applied to claims 1 and 7 above, and in further view of Scibor et al. Pub. No. US 20210103422 A1 (hereinafter Scibor).
Regarding claim 8, the combination of Li and Choi teaches all of the elements of claim 7 as shown in the rejection above, but does not explicitly teach: wherein the stop condition comprises: monitoring, with the autoencoder architecture, a history of label probability purity values associated with the processed each unlabeled data item not increasing over a threshold number of iterations of processing unlabeled data items by the autoencoder architecture.
However, in the analogous art, Scibor teaches: wherein the stop condition comprises: monitoring, with the autoencoder architecture, a history of label probability purity values associated with the processed each unlabeled data item not increasing over a threshold number of iterations of processing unlabeled data items by the autoencoder architecture ([0087] teaches utilizing an early stopping heuristic that requires a positive change in the validation accuracy (i.e., indicative of a label probability purity value) within ten epochs (i.e., a threshold number of iterations)).
It would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to have modified the combination of Li and Choi with the above teaching of Scibor because doing so would lead to an expected reduction in computation time (Scibor, [0087]).
Regarding claim 9, the combination of Li and Choi teaches all of the elements of claim 1 as shown in the rejection above, but does not explicitly teach: wherein the stop condition comprises: monitoring, with the autoencoder architecture, a history of label probability purity values associated with the processed each unlabeled data item decreasing over one or more iterations of processing unlabeled data items by the autoencoder architecture.
However, in the analogous art, Scibor teaches: wherein the stop condition comprises: monitoring, with the autoencoder architecture, a history of label probability purity values associated with the processed each unlabeled data item decreasing over one or more iterations of processing unlabeled data items by the autoencoder architecture ([0087] teaches utilizing an early stopping heuristic that requires a positive change in the validation accuracy (i.e., indicative of a label probability purity value) within ten epochs (i.e., over one or more iterations)).
It would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to have modified the combination of Li and Choi with the above teaching of Scibor because doing so would lead to an expected reduction in computation time (Scibor, [0087]).
Regarding claim 10, the combination of Li, Choi, and Scibor teaches all of the elements of claim 9 as shown in the rejection above. Scibor further teaches: wherein the stop condition comprises: monitoring, with the autoenc