DETAILED ACTION
Claims 1-15 are pending in the Instant Application.
Claims 1-15 are rejected (Non-Final Rejection).
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 .
Priority
The Instant Application, filed 08/30/2024, claims foreign priority to
10-2023-0114709 (KR), filed 08/30/2023. Thus, the effective filing date for what is recited therein is 08/30/2023.
Information Disclosure Statement
The information disclosure statements (IDSs) submitted on 05/12/2026 and 04/28/26 were considered by the examiner.
Claim Objections
Claim 4 is objected to because of the following informalities: a typographical error. Claim 4 recites a DMS-5 depressive disorder diagnostic criteria. The actual term is DSM-5. Appropriate correction is required.
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 8-15 are rejected under 35 U.S.C. 101 because the claimed invention is directed to non-statutory subject matter. The claims do not fall within at least one of the four categories of patent eligible subject matter because claim 8 recites a “computer program,” which is non-statutory computer software per se, capable of protection under copyright, but not by patent. Claims 1-7 perform a method, and thus any version of claim 8 is still considered non-statutory.
Claim 9 recites “A training device,” but does not explicitly recite any hardware. The specification at [0078] explains that the invention can be implemented as software or hardware. Software would be considered non-statutory subject matter. Please, limit the device in such a way that the claims cannot be considered software to overcome this rejection. Thus, claim 9 is rejected. Claims 10-15 are dependent on claim 9, do not cure the issue and are rejected for the same rational and reasoning.
Claim Rejections - 35 USC § 103
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
Claims 1-5 and 9-13 are rejected under 35 U.S.C. 103 as being unpatentable over Farruque, “Depression Symptoms Modelling from Social Media Text: An Active Learning Approach,” in view of Xiao et al. (“Xiao”), United States Patent No. 11,955,240.
As per claim 1, Farruque discloses a training method comprising:
performing transfer learning a first artificial neural network model based on a dataset ([Page 4, 3.1 Clinician annotated dataset curation] and [Page 10, 5.1 Active Learning (AL) framework) wherein transfer learning is performed using human annotated depression tweets) on a pre-trained Mental-BERT);
inputting depressive disorder-related expression data into the first artificial neural network model, labeling the depressive disorder-related expression data according to depressive disorder diagnosis criteria ([Page 6, 3.2 Annotation Task Description] and [Page 10, 5.1.1 Step 1:Creating first DSD model] wherein depressive disorder-related tweets are input into the neural network to generate labels according to diagnosis criteria (symptoms)), and generating labeled depressive disorder-related expression data ([Page 11, 5.1.2 Step 2: Harvesting tweets using DSD-Clinician-1] wherein a labeled data set is generated of depressive disorder-related expression data); and training a second artificial neural network model based on the labeled depressive disorder-related expression data so that a second artificial neural network model may output the depressive disorder diagnosis criteria corresponding to input data ([Page 13, 5.1.4 Step 4: Creating a second DSD Model] wherein a second model, “DSD-Clinician-2” is trained based on labeled data (See [Page 14, Table 13], which outputs diagnosis criteria (symptom categories) from input text), but does not disclose where the dataset is a consultation dataset sentence. However, Xiao teaches where the dataset is a consultation dataset sentence ([Col 10, lines 49-67] wherein a consultation dataset is described, labeled for disease).
Farruque discloses fine-tuning pretrained Mental-BERT to create DSD clinician-1 (the first neural network model) using annotated tweets, but does not expressly use training examples from consultation sentences. However, Xiao uses consultation sentences. One could use the consultation sentence from Xiao instead of the annotated tweets from Farruque to teach the claimed invention. It would have been obvious to one of ordinary skill in the art to combine the method of generating training data to train a model to output depressive disorder diagnosis criteria in Farruque with the use of consultation sentences in Xiao to provide greater context and improve accuracy.
As per claim 2, note the rejection of claim 1 where Farruque and Xiao are combined. The combination teaches the training method of claim 1. Farruque further discloses wherein the generating of labeled depressive disorder-related expression data comprises: extracting a plurality of depressive disorder diagnostic criteria corresponding to the depressive disorder-related expression data from the depressive disorder diagnostic criteria ([Page 6, 3.2 Annotation Task Description] wherein labeled data is generated by extracting 10 diagnostic criteria); and
multi-labeling the depressive disorder-related expression data based on the plurality of depressive disorder diagnostic criteria ([Page 6, All Sections] wherein the tweets are labeled using multiple labels based on the criteria).
As per claim 3, note the rejection of claim 1 where Farruque and Xiao are combined. The combination teaches the training method of claim 1. Farruque further discloses wherein the generating of labeled depressive disorder-related expression data comprises:
obtaining first output data by inputting the depressive disorder-related expression data into the first artificial neural network model ([Pages 6 and 7, 3.4 Depression Symptoms Annotation Process] wherein a first output data consisting of a label is obtained);
converting the first output data into second output data based on a sigmoid function, the second output data comprising a plurality of components corresponding to the depressive disorder diagnostic criteria, respectively ([Appendix B] wherein the output comprising a value for each of the node labels, is converted based on a sigmoid function); and labeling the depressive disorder-related expression data based on the depressive disorder diagnostic criteria corresponding to a component equal to or greater than a preset threshold value from among the plurality of components ([Page 12, 5.1.3 Step 3: Harvesting tweets using best ZSL Model] wherein a threshold is used to choose the label using the converted score).
As per claim 4, note the rejection of claim 1 where Farruque and Xiao are combined. The combination teaches the training method of claim 1. Farruque further discloses wherein the depressive disorder diagnostic criteria comprise DMS-5 depressive disorder diagnostic criteria. ([Page 6, 3.3 Annotation Guideline Creation] wherein DSM-5 is used as the base reference).
As per claim 5, note the rejection of claim 1 where Farruque and Xiao are combined. The combination teaches the training method of claim 1. Farruque further discloses wherein the training comprises: training the second artificial neural network model so that the second artificial neural network model may output the depressive disorder diagnosis criteria corresponding to the input data and probability corresponding to the depressive disorder diagnosis criteria ([Page 14] wherein the model DSD-Clinician 2 is trained to output the depressive disorder diagnosis criteria (symptom in the prior art), along with the [Page 23, Appendix B] wherein the sigmoid function provides the output probability as described).
Aas per claim 9, Farruque discloses a training device comprising: a first artificial neural network model that performs transfer learning on sentences from a dataset labels depressive disorder-related expression data according to depressive disorder diagnosis criteria ([Page 4, 3.1 Clinician annotated dataset curation] and [Page 10, 5.1 Active Learning (AL) framework) wherein transfer learning is performed using human annotated depression tweets) on a pre-trained Mental-BERT), and generates labeled depressive disorder-related expression data ([Page 11, 5.1.2 Step 2: Harvesting tweets using DSD-Clinician-1] wherein a labeled data set is generated of depressive disorder-related expression data); and a second artificial neural network model trained to output the depressive disorder diagnosis criteria corresponding to the labeled depressive disorder-related expression data ([Page 13, 5.1.4 Step 4: Creating a second DSD Model] wherein a second model, “DSD-Clinician-2” is trained based on labeled data (See [Page 14, Table 13], which outputs diagnosis criteria (symptom categories) from input text), but does not disclose where the dataset is a consultation dataset sentence. However, Xiao teaches where the dataset is a consultation dataset sentence ([Col 10, lines 49-67] wherein a consultation dataset is described, labeled for disease).
Farruque discloses fine-tuning pretrained Mental-BERT to create DSD clinician-1 (the first neural network model) using annotated tweets, but does not expressly use training examples from consultation sentences. However, Xiao uses consultation sentences. One could use the consultation sentence from Xiao instead of the annotated tweets from Farruque to teach the claimed invention. It would have been obvious to one of ordinary skill in the art to combine the method of generating training data to train a model to output depressive disorder diagnosis criteria in Farruque with the use of consultation sentences in Xiao to provide greater context and improve accuracy.
As per claim 10, note the rejection of claim 9 where Farruque and Xiao are combined. The combination teaches the training device of claim 9. Farruque further discloses wherein the first artificial neural network model is configured to: extract a plurality of depressive disorder diagnostic criteria corresponding to the depressive disorder-related expression data from the depressive disorder diagnostic criteria ([Page 6, 3.2 Annotation Task Description] wherein labeled data is generated by extracting 10 diagnostic criteria); and multi-label the depressive disorder-related expression data based on the plurality of depressive disorder diagnosis criteria to generate the labeled depressive disorder-related expression data ([Page 6, All Sections] wherein the tweets are labeled using multiple labels based on the criteria).
As per claim 11, note the rejection of claim 9 where Farruque and Xiao are combined. The combination teaches the training device of claim 9. Farruque further discloses wherein the first artificial neural network model is configured to: obtain first output data by inputting depressive disorder-related expression data into the first artificial neural network model([Pages 6 and 7, 3.4 Depression Symptoms Annotation Process] wherein a first output data consisting of a label is obtained); convert the first output data into second output data including a plurality of components corresponding to the depressive disorder diagnosis criteria, respectively, based on a sigmoid function ([Appendix B] wherein the output comprising a value for each of the node labels, is converted based on a sigmoid function); and label the depressive disorder-related expression data based on the depressive disorder diagnostic criteria corresponding to a component equal to or greater than a preset threshold value from among the plurality of components to generate the labeled depressive disorder-related expression data ([Page 12, 5.1.3 Step 3: Harvesting tweets using best ZSL Model] wherein a threshold is used to choose the label using the converted score).
As per claim 12, note the rejection of claim 9 where Farruque and Xiao are combined. The combination teaches the training device of claim 9. Farruque further discloses wherein the depressive disorder diagnostic criteria comprise DMS-5 depressive disorder diagnostic criteria ([Page 6, 3.3 Annotation Guideline Creation] wherein DSM-5 is used as the base reference).
As per claim 13, note the rejection of claim 9 where Farruque and Xiao are combined. The combination teaches the training device of claim 9. Farruque further discloses wherein the second artificial neural network model is trained to output the depressive disorder diagnosis criteria corresponding to the labeled depressive disorder-related expression data and probability corresponding to the depressive disorder diagnosis criteria ([Page 14] wherein the model DSD-Clinician 2 is trained to output the depressive disorder diagnosis criteria (symptom in the prior art), along with the [Page 23, Appendix B] wherein the sigmoid function provides the output probability as described).
.Claims 7 and 15 are rejected under 35 U.S.C. 103 as being unpatentable over Farruque, in view of Xiao, in further view of KIM et al. (“Kim”), KR-20230014034-A (Machine translated copy provided).
As per claim 7, note the rejection of claim 1 where Farruque and Xiao are combined. The combination teaches the method of claim 1, but does not disclose wherein the first artificial neural network model comprises a KoBERT model, and the second artificial neural network model comprises a GRU model. However, Kim teaches wherein the first artificial neural network model comprises a KoBERT model ([Page 11, Last paragraph] wherein a KoBERT model is described being used), and the second artificial neural network model comprises a GRU model ([Page 3, Last paragraph] wherein a Gated Recurrent Unit model is shown to be excellent with sentence data where the previous word is reflected in the next word). The Supreme Court in KSR Int'l Co. v. Teleflex Inc., 550 U.S. 398, 415-421, 82 USPQ2d 1385, 1395-97 (2007) identified a number of rationales to support a conclusion of obviousness which are consistent with the proper “functional approach” to the determination of obviousness as laid down in Graham. One such rationale is (B) Simple substitution of one known element for another to obtain predictable results. In this case the combination would be obvious because substitution of one model in a spoken language with the same model trained on another language would obtain predictable results since the only difference would be a better model specific to a particular language.
As per claim 15, claim 15 is the device that performs the method of claim 7 and is rejected for the same rationale.
Claims 6 and 14 are rejected under 35 U.S.C. 103 as being unpatentable over Farruque, in view of Xiao, in further view Chen et al. (“Chen”), “Semi-Supervised Models via Data Augmentation for Classifying Interactive Affective Responses.”
As per claim 6, note the rejection of claim 1 where Farruque and Xiao are combined. The combination teaches the method of claim 1, Farruque further discloses residual data from among the depressive disorder-related expression data that is not labeled by the first artificial neural network model ([Page 11, 5.1.2. Harvesting tweets using DSD-Clinician-1] wherein some samples have labels without enough accuracy to be labeled);but does not disclose performing data augmentation; and re-inputting residual data expanded by the data augmentation into the first artificial neural network to label the expanded residual data according to the depressive disorder diagnosis criteria. However, Chen teaches performing data augmentation ([Page 6, 4.2 Unsupervised Learning, “Paraphrasing Unlabeled Sentences”] wherein back translations provide an augmented data set); and re-inputting residual data expanded by the data augmentation into the first artificial neural network to label the expanded residual data ([Page 7, 4.2 Unsupervised Learning, “Consistency REgulatiztion for Agumented Sentences] wherein equation 6 describes the model before and after applying the larger augmented data set).
Farrague discloses labeling depressive disorder-related expression data, but does not describe augmentation and the re-inputting of that data into the model. One could apply the augmentation and re-inputting of data to the depressive disorder-related expression data as in Farrague to teach the claimed invention. It would have been obvious to one of ordinary skill in the art before the effective filing date of the invention to combine the method of using a first model to generate input to a second model using consultation sentences in the combination of Farruque and Xiao with the augmentation in Chen in order to increase useful training data and provide better classification consistency.
As per claim 14, claim 14 is the device that performs the method of claim 6 and is rejected for the same rationale.
Claim 8/1-8/7 are rejected under 35 U.S.C. 103 as being unpatentable over the respective combination of references applied to claims 1-7 as shown below. (The expression 8/n, indicates the claim including both the limitations from claim 8 and claim n, wherein n can be 1-7.)
Claims 8/1, 8/2, 8/3, 8/4 and 8/5 are all rejected by the combination of Farruque and Xiao, each rejected for the same rationale and reasoning as the rejection of claim n.
Claim 8/6 is rejected by the combination of Farruque, Xiao and Chen, rejected for the same rationale and reasoning as the rejection of claim 6.
Claim 8/7 is rejected by the combination of Farruque, Xiao and Kimm rejected for the same rationale and reasoning as the rejection of claim 7.
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to KANNAN SHANMUGASUNDARAM whose telephone number is (571)270-7763. The examiner can normally be reached M-F 9:00 AM -6:00 PM.
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, Charles Rones can be reached at (571) 272-4085. 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.
/KANNAN SHANMUGASUNDARAM/Primary Examiner, Art Unit 2168
Prosecution Insights