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 notes the entry of the following papers:
Amended claims filed 12/11/2025.
Applicant arguments/remarks made in amendment filed 12/11/2025.
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and the fee set forth in 37 CFR 1.17(e) has been timely paid, the finality of the previous Office action has been withdrawn pursuant to 37 CFR 1.114.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 1/12/2026 has been entered.
Claims 1, 8 and 15 are amended. Claims 1-20 are presented for examination.
Response to Arguments
Applicant presents arguments. Each is addressed.
Applicant argues “None of the problems and solutions as provided in amended claim one or its features as provided below are used.” (Remarks, page 8, paragraph 1, line 5.) Applicant then recites the entirety of the active limitations of the claim and argues that Chen, Gallant, and Meera do not cure the deficiencies of Ampanavos without specifically identifying which limitations each of the cited art fails to teach. The arguments are moot in view of new grounds of rejection necessitated by amendment.
Applicant argues “Independent claims 8 and 15 provide similar language.” (Remarks, page 9, paragraph 2, line 1.) However, claim 1 remains rejected. Claims 8 and 15, which provide similar language, are similarly rejected.
Applicant argues “The other claims are dependent from claims one of the independent claims 1, 11, and 20 as discussed above, and are therefore believed patentable for at least the same reasons.” (Remarks, page 9, paragraph 4, line 1.) However, the independent claims remain rejected. The dependent claims remain rejected at least for depending from rejected base claims.
Subject Matter Eligibility
In determining whether the claims are subject matter eligible, the examiner has considered and applied the 2019 USPTO Patent Eligibility Guidelines, as well as guidance in the MPEP chapter 2106. The examiner finds that the independent claims do not recite a judicial exception.
Claim Rejections - 35 USC § 103
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis (i.e., changing from AIA to pre-AIA ) for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
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 1-20 are rejected under 35 U.S.C. § 103 as being unpatentable over Ampanavos, et al (Structural Design Recommendations in the Early Design Phase using Machine Learning, herein Ampanavos), Gallant, S., et al (Image retrieval using Image Context Vectors: first results, herein Gallant), Chen, et al (Research on Recommendation Method of Product Design Scheme Based on Multi-Way Tree and Learning-to-Rank, herein Chen), Meera, et al (Object Recognition in Images, herein Meera), and Durand, et al (A Learning Design Recommendation System Based on Markov Decision Processes, herein Durand).
Regarding claim 1,
Ampanavos teaches a method, by a processor, for providing enhanced generative models based assistance in a computing environment (Ampanavos, Figs. 1 and 3, and page 3, paragraph 4, line 2 “we are proposing a method for structural design approximation that directly uses sketch-level plans.” And, page 3, paragraph 5, line 1 “ApproxiFramer aims to inform the early phase design exploration in a sketch-based environment through the real-time generation of structural designs. Figure 1 describes how ApproxiFramer can be integrated into such an environment. A user-generated sketch first passes through a pre-processing step that converts a noisy and imprecise input to a clean drawing so that it can be used with our predictive system.”
PNG
media_image1.png
369
793
media_image1.png
Greyscale
PNG
media_image2.png
352
789
media_image2.png
Greyscale
Examiner notes that one of ordinary skill in the art would understand the primary reference as teaching execution of a computer (processor) implemented method. “[I]n considering the disclosure of a reference, it is proper to take into account not only specific teachings of the reference but also the inferences which one skilled in the art would reasonably be expected to draw therefrom.” MPEP § 2144.01. In other words, method is method, processing step is by a processor, and ApproxiFramer is enhanced generative model providing assistance.), comprising:
receiving a partially completed design of an object (Ampanavos, Fig. 1, In other words, input is receiving, and sketch (a) is partially completed design of an object.) ;
taking at least a partial image of said partially completed design and generating a [context] vector by a [context] analyzer (Ampanavos, Fig. 2, and, page 5, paragraph 2, line 1 “The input of the system is in image format, providing significant flexibility to the user in the selection of design software or medium. In the core of the system lies a convolutional neural net (CNN) that we trained to take an image of a sketch, representing a plan of a building layout, and predict the position of a group of columns.” And, page 8, paragraph 3, line 1 “The model successfully identified when to stop adding new columns 100% of the time. During training, we used a zero vector to indicate the stopping point of the predictions. During inference, we modified the threshold to be a vector where at least one of the x or y coordinates has a value less than 2.”
PNG
media_image3.png
655
793
media_image3.png
Greyscale
In other words, image is one or more partial images, vector is vector of the one or more partial images, column vector is vector represents at least a portion of the object.) ;
processing said [context] vector so as to determine one or more objects representative of said at least partial image and a location representing a plurality of coordinates corresponding to an object to be added (Ampanavos, Fig. 1, and, page 193, paragraph 4, line 4 “As such, the structures that we design can be easily abstracted to a diagrammatic level. A set of coordinates that indicate the locations of the columns is then a sufficient description of such a frame. The developed system generates structural layouts for orthogonal, sketch-level, single-floor plans. These plans include exterior and interior walls of the building, both represented by single straight lines that are either horizontal or vertical.”
In other words, from prior mapping, image of a sketch is partial image, vector is vector, columns are one or more objects, and, set of coordinates that indicate locations of the columns is a location representing a plurality of coordinates corresponding to an object to be added.) ;
[determining using said context vector a missing concept shadow wherein the vector is calculated to have a texture and a location];
generating a [set of recommendations] for completing the partially completed design based on one or more generative models (Ampanavos, Fig. 1, In other words, outputs solution (d) is generating a completed design, and using ApproxiFramer is based on one or more generative models.)
using a design analyzer having at least [an object recognition component] (Ampanavos, page 3, paragraph 5, line 1 “ApproxiFramer aims to inform the early phase design exploration in a sketch-based environment through the real-time generation of structural designs.” In other words, ApproxiFramer is a design analyzer.) and
[a set of difference method to be executed with respect to a closest matching image in a representative dataset], wherein said set of recommendations includes
[one or more calculations using statistical modeling using statistical distributions to enhance said partial completed design].
Thus far, Ampanavos does not explicitly teach “context” vector.
Gallant teaches context vector (Gallant, Figure 1, and, page 82, paragraph 2, line 1 “A context vector is a high (-300) dimensional vector that can represent images, sub-images, or image queries. Image context vectors are an extension of previous work in document retrieval where context vectors were used to represent documents, terms, and queries.”
PNG
media_image4.png
701
706
media_image4.png
Greyscale
In other words, context vector is context vector.)
Gallant teaches determining using said context vector a missing concept shadow wherein the vector is calculated to have a texture and a location (Gallant, Page 85, paragraph 1, line 1 “2.1 Creation of image context vectors. The first processing step is to use image feature detectors to extract between 10 and 300 key features and their corresponding locations in the image. Features might include angles, homogeneous areas (and their centroids), shapes, textures, etc.” Examiner notes the specification of the instant application recites “The context vector may also represent a missing concept such as, for example, a shadow in which case the vector would look be a pair that includes a texture and location…” Based on this, examiner is interpreting that a concept, such as a shadow would be a texture, location pair encoded in the vector. In other words, context vector is context vector, features include textures is textures, and features and their corresponding locations is texture location pair.)
Gallant teaches a set of difference method to be executed with respect to a closest matching image in a representative dataset (Gallant, page 93, paragraph 2, line 1 “For retrieval, all 2200 images (or, optionally, the 86 full images) are ranked by closeness to
the query image context vector. Searching the entire database of images takes only about 3 seconds.” In other words, entire database is representative dataset, ranked by closeness is difference method, and ranked by closeness to the query image context vector is closest matching image.)
Both Ampanavos and Gallant are directed to processing images, among other things. Ampanavos teaches a method, by a processor, for providing enhanced generative models based assistance in a computing environment, comprising receiving a partially completed design of an object, taking at least a partial image of said partially completed design and generating a vector by an analyzer, processing said vector so as to determine one or more object representative of said at least partial image and a location representing a plurality of coordinates corresponding to an object to be added; but does not explicitly teach the vector is a context vector or determining, using the context vector, a missing concept wherein the vector is calculated to have a texture and a location, and a difference method to be executed with respect to a closest matching image in a representative dataset. Gallant teaches the vector is a context vector and determining, using the context vector, a missing concept wherein the vector is calculated to have a texture and a location and a difference method to be executed with respect to a closest matching image in a representative dataset.
In view of the teaching of Ampanavos, it would be obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine the teaching of Gallant into Ampanavos. This would result in a method, by a processor, for providing enhanced generative models based assistance in a computing environment, comprising receiving a partially completed design of an object, taking at least a partial image of said partially completed design and generating a vector by an analyzer, processing said vector so as to determine one or more object representative of said at least partial image and a location representing a plurality of coordinates corresponding to an object to be added where the vector is a context vector, and determining, using the context vector, a missing concept wherein the vector is calculated to have a texture and a location and a difference method to be executed with respect to a closest matching image in a representative dataset.
One of ordinary skill in the art would be motivated to do this in order to improve the speed and accuracy of searching image databases. (Gallant, page 83, paragraph 1 “Image databases are rapidly proliferating, and a growing segment of our economy is devoted to producing video imagery for private consumption. In the Government domain, the recent Magellan mission to Venus returned 30,000 1 Mbyte radar images of the planet, more data than that from all previous NASA interplanetary missions combined1. Effective and rapid search of image databases is becoming an increasingly acute problem. The most common approach is to hand-label images with text, and then to perform keyword searches on the text alone. However exponential growth in image data requires the ability to automatically process images into efficiently searchable representations for later querying. Moreover, manual indexing of text is notoriously inconsistent and error prone, and there is little reason to believe that indexing of pictures produces better results.”)
Thus far, the combination of Ampanavos and Gallant does not explicitly teach generating a set of recommendations.
Chen teaches generating a set of recommendations (Chen, abstract, line 3 “In order to improve the efficiency of product design, a product design scheme recommendation algorithm based on multi-way tree and learning-to-rank is proposed.” And, page 8, paragraph 2, line 1 “Given a set of product suppliers and serial numbers S = {S1, S2, …, Sn} describing these products’ attributes’ combination based on the designer’s preference, our goal was to induce a function f that scored design schemes in S according to the degree to which they satisfy the specific requirements of a designer, given as an n-tuple (s1, s2, … , sn). In other words, design schemes are set of recommendations.)
Both Chen and the combination of Ampanavos and Gallant are directed to using machine learning to generate design recommendations, among other things. The combination of Ampanavos and Gallant teaches a method, by a processor, for providing enhanced generative models based assistance in a computing environment, comprising receiving a partially completed design of an object, taking at least a partial image of said partially completed design and generating a context vector by a context analyzer, processing said context vector so as to determine one or more object representative of said at least partial image and a location representing a plurality of coordinates corresponding to an object to be added, determining using said context vector a missing concept shadow wherein the vector is calculated to have a texture and a location; but does not explicitly teach generating a set of recommendations. Chen teaches generating a set of recommendations.
In view of the teaching of the combination of Ampanavos and Gallant, it would be obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine the teaching of Chen into the combination of Ampanavos and Gallant. This would result in a method, by a processor, for providing enhanced generative models based assistance in a computing environment, comprising receiving a partially completed design of an object, taking at least a partial image of said partially completed design and generating a context vector by a context analyzer, processing said context vector so as to determine one or more object representative of said at least partial image and a location representing a plurality of coordinates corresponding to an object to be added, determining using said context vector a missing concept shadow wherein the vector is calculated to have a texture and a location, and generating a set of recommendations for completing the design.
One of ordinary skill in the art would be motivated to do this in order to improve design efficiency. (Chen, abstract, line 1 “A product is composed of several components, and the number, type, and combination of components plays a crucial role in the process of product design. It is difficult to get an optimized scheme in a short time. In order to improve the efficiency of product design, a product design scheme recommendation algorithm based on multi-way tree and learning-to-rank is proposed. Firstly, the product solution model, whose nodes are obtained by mapping the product attributes, is generated according to the design process, and the alternative scheme is obtained by traversing the multi-tree model.”)
Thus far, the combination of Ampanavos, Gallant, and Chen does not explicitly teach an object recognition component.
Meera teaches an object recognition component (Meera, abstract, line 3 “Object recognition have immense of applications in the field of monitoring and surveillance, medical analysis, robot localization and navigation etc. The appearance of an object can be varied due to scene clutter, photometric effects, changes in shape and viewpoints of the object. The recognition should be invariant to viewpoint changes and object transformations, robust to noise and occlusion. This work aims at formulating a technique that consist of two stages. For the first stage, the query image is categorized using a classifier. For classifier optimization we have implemented two types of classifiers-Support Vector Machine(SVM) classifier that make use of GIST features and k-nearest neighbour (kNN) classifier that make use of Scale Invariant Feature Transform(SIFT). ” In other words, object recognition is object recognition, and classifier is object recognition component. Examiner notes that the previously mapped method “ApproxiFramer”, is a design analyzer.)
Both Meera and the combination of Ampanavos, Gallant, and Chen are directed to images and objects within images, among other things. The combination of Ampanavos and Gallant teaches a method, by a processor, for providing enhanced generative models based assistance in a computing environment, comprising receiving a partially completed design of an object, taking at least a partial image of said partially completed design and generating a context vector by a context analyzer, processing said context vector so as to determine one or more object representative of said at least partial image and a location representing a plurality of coordinates corresponding to an object to be added, determining using said context vector a missing concept shadow wherein the vector is calculated to have a texture and a location, and generating a set of recommendations for completing the design; but does not explicitly teach using an object recognition component. Meera teaches using an object recognition component.
In view of the teaching of Ampanavos, Gallant, and Chen it would be obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine the teaching of Meera into the combination of Ampanavos, Gallant, and Chen. This would result in a method, by a processor, for providing enhanced generative models based assistance in a computing environment, comprising receiving a partially completed design of an object, taking at least a partial image of said partially completed design and generating a context vector by a context analyzer, processing said context vector so as to determine one or more object representative of said at least partial image and a location representing a plurality of coordinates corresponding to an object to be added, determining using said context vector a missing concept shadow wherein the vector is calculated to have a texture and a location, and generating a set of recommendations for completing the design using an object recognition component.
One of ordinary skill in the art would be motivated to do this because object recognition has many important applications, and an efficient object recognition method would be valuable. (Meera, page 126, column 1, paragraph 2, line 5 “Object recognition has many applications in the field of biometric recognition, surveillance, industrial inspection, content based image retrieval(CBIR), robotics, medical analysis, human-computer interaction, intelligent vehicle systems etc. [8]. Object recognition becomes tedious due to the positioning, scaling, alignment and occlusion of objects. Indoor and outdoor images for same object can have varying lighting condition. Occlusion is the condition when an object is not fully visible in an image. An efficient object recognition method should be capable of handling such difficulties. The key objective is to develop object recognition techniques which are efficient and less complex.”)
Thus far, the combination of Ampanavos, Gallant, Chen, and Meera does not explicitly teach one or more calculations using statistical modeling using statistical distributions to enhance said partial completed design.
Durand teaches one or more calculations using statistical modeling using statistical distributions to enhance said partial completed design (Durand, page 4, paragraph 3, line 1 “ILD-RS (intelligent learning design recommendation system) is a software component able to recommend to the learning designer appropriate learning paths during the learning design phase: ILD-RS does not substitute but assist the learning designer.” And, page 4, paragraph 4, line 4 “By using MDP, the objective is to simulate some of the features of the teacher’s decision process in order to offer learning path recommendations which teachers are free to accept or decline while they are building their own learning activities.
A Markov decision process is a 4-tuple where:
S is a finite set of states
A is a finite set of actions
T:S X A -> Prob(s) specifies a probability distribution for each state and action over next states
R(s,a) is a reward function.
A MDP is assumed to respect the Markov property; the probability distribution T is constituted of independent probabilities.” Examiner notes that the specification of the instant application recites “Additionally, the enhanced design model assistant service 410 (using one or more components therein) may perform one or more various types of calculations or computations.” (Specification, paragraph [0071], line 1.) The design model assistant is performing the calculation or computation. There is no support in the specification for “recommending” the performance of a calculation or a type of statistical modeling. Based on this, Examiner is interpreting that the calculations are performed in determining the recommendation(s). In other words, using MDP (Markov decision process) is using statistical modeling, and probability distribution T is using statistical distribution.)
Both Durand and the combination of Ampanavos, Gallant, Chen, and Meera are directed to making design recommendations, among other things. The combination of Ampanavos, Gallant, Chen, and Meera teaches a method, by a processor, for providing enhanced generative models based assistance in a computing environment, comprising receiving a partially completed design of an object, taking at least a partial image of said partially completed design and generating a context vector by a context analyzer, processing said context vector so as to determine one or more object representative of said at least partial image and a location representing a plurality of coordinates corresponding to an object to be added, determining using said context vector a missing concept shadow wherein the vector is calculated to have a texture and a location, and generating a set of recommendations for completing the design using an object recognition component; but does not explicitly teach one or more calculations using statistical modeling using statistical distributions to enhance said partial completed design. Durand teaches one or more calculations using statistical modeling using statistical distributions to enhance said partial completed design.
In view of the teaching of the combination of Ampanavos, Gallant, Chen, and Meera, it would be obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine Durand into the combination of Ampanavos, Gallant, Chen, and Meera. This would result in a method, by a processor, for providing enhanced generative models based assistance in a computing environment, comprising receiving a partially completed design of an object, taking at least a partial image of said partially completed design and generating a context vector by a context analyzer, processing said context vector so as to determine one or more object representative of said at least partial image and a location representing a plurality of coordinates corresponding to an object to be added, determining using said context vector a missing concept shadow wherein the vector is calculated to have a texture and a location, and generating a set of recommendations for completing the design using an object recognition component, and one or more calculations using statistical modeling using statistical distributions to enhance said partial completed design.
One of ordinary skill in the art would be motivated to do this in order to assist designers by reducing work overload. (Durand, page 1, paragraph 4, line 1 “Cognitive overload had already been identified as a big challenge for Learning Management Systems’ editors since the recent integration of numerous business features that significantly raised the amount of information processing required to use these tools. Considering this state of affairs, the main actors of the e-learning industry, though initially interested, did not take up the proposed learning design approaches as they preferred to find other solutions suitable for helping teachers, but at the same time minimizing the cognitive overload.”)
Regarding claim 2,
The combination of Ampanavos, Gallant, Chen, Meera, and Durand teaches the method of claim 1, where
further including identifying one or more sections of the partially completed design for applying the set of recommendations (Chen, abstract, line 3 discloses applying set of recommendations. Ampanavos, Fig. 1. discloses identifying one or more sections of the partially completed design. In other words, the input sketch is one or more sections of a partially completed design for applying the set of recommendations.).
Regarding claim 3,
The combination of Ampanavos, Gallant, Chen, Meera, and Durand teaches the method of claim 1, further including
generating one or more context vectors from one or more partial images of the partially completed design, wherein a context vector represents at least a portion of the object (Ampanavos, Fig. 2, and, page 5, paragraph 2, line 1 “The input of the system is in image format, providing significant flexibility to the user in the selection of design software or medium. In the core of the system lies a convolutional neural net (CNN) that we trained to take an image of a sketch, representing a plan of a building layout, and predict the position of a group of columns.” And, page 8, paragraph 3, line 1 “The model successfully identified when to stop adding new columns 100% of the time. During training, we used a zero vector to indicate the stopping point of the predictions. During inference, we modified the threshold to be a vector where at least one of the x or y coordinates has a value less than 2.” In other words, image is one or more partial images, vector is context vector of the one or more partial images, column vector is vector represents at least a portion of the object.)
Regarding claim 4,
The combination of Ampanavos, Gallant, Chen, Meera, and Durand teaches the method of claim 1, further including
generating a list of candidate suggestions for adding to the set of recommendations (Chen, page 8, paragraph 1, line 1 “The list of solutions generated based on AHP-TOPSIS did not reflect the designer’s preference for the component supplier combination. In order to solve this problem, a component supplier combination recommendation algorithm based on user behavior feedback was proposed to further optimize the recommendation list.” In other words, list of solutions is list of candidate suggestions.).
Regarding claim 5,
The combination of Ampanavos, Gallant, Chen, Meera, and Durand teaches the method of claim 1, further including
ranking each suggestion in a list of candidate suggestions adding to the set of recommendations (Chen, page 6, paragraph 1, line 9 “The basic idea of TOPSIS is to determine the ideal solution and the negative ideal solution, according to the elements of the decision-making problem, and to rank each scheme according to the relative closeness degree of the ideal solution of its elements.” In other words, rank each scheme is ranking each suggestion in a list of suggestions.).
Regarding claim 6,
The combination of Ampanavos, Gallant, Chen, Meera, and Durand teaches the method of claim 1, further including
generating one or more enhanced images of the images from one or more incomplete images and a context vector using a generative model (Ampanavos, Algorithm 1.
PNG
media_image5.png
263
833
media_image5.png
Greyscale
In other words, predict next column and append to image is generating one or more enhanced images from one or more incomplete images, and column is vector.)
Regarding claim 7,
The combination of Ampanavos, Gallant, Chen, Meera, and Durand teaches the method of claim 1, further including initializing a machine learning mechanism to:
learn the partially completed design of the object (Ampanavos, Fig. 1, In other words, system converts a noisy sketch to a clean drawing is learn the partially completed design of the object.) ;
train a machine learning model to learn and identify set of recommendations (for completing the partially completed design based on one or more generative models (Chen page 6, paragraph 1, line 9 discloses a set of recommendations, (see office action, page 8.) Ampanavos, page 5, paragraph 2, line 1 “The input of the system is in image format, providing significant flexibility to the user in the selection of design software or medium. In the core of the system lies a convolutional neural net (CNN) that we trained to take an image of a sketch, representing a plan of a building layout, and predict the position of a group of columns. In each iteration, the newly predicted columns are added to the solution, and a new image is rendered that contains both the initial sketch and the columns that have been placed so far. This newly rendered image is then used as the input of the next iteration.” In other words, CNN is machine learning model, trained is train, image of a sketch is partially completed design, and new image is completed design.); and
automatically select or modify the set of recommendations (Chen, page 8, step 8 “Step 8: Sort each design scheme according to the size order of Di, and select the first n1 schemes for the designer to choose.” In other words, select the first n1 schemes is automatically select the set of recommendations.) for
completing the partially completed design based on one or more generative models and collected feedback (Ampanavos, Fig. 1. And, abstract, line 18 “We trained a Convolutional Neural Net to iteratively generate structural design solutions for sketch-level building plans using a synthetic dataset and achieved an average error of 2.2% in the predicted positions of the columns.” In other words, sketch is partially completed design, ApproxiFramer is one or more generative model, structural solution is completing the design, and iterative design implies collected feedback.).
Claims 8-14 are systems comprising one or more computers with executable instructions that when executed by the system claims corresponding to method claims 1-7, respectively. Otherwise, they are the same. The combination of Ampanavos, Gallant, Chen, Meera, and Durand teaches a system (Ampanavos, page 2, paragraph 2, line 1 “In this paper, we introduce ApproxiFramer, an automated system with the ability to generate structural design recommendations during the conceptual phase of architectural design.” In other words, automated system is a system comprising one or more computers with executable instructions.) Therefore, claims 8-14 are rejected for the same reasons as claims 1-7.
Claims 15-17, and 19-20 are one or more computer readable storage media and program instructions stored on the one or more computer readable storage media claims corresponding to method claims 1-3, and 6-7, respectively. Otherwise, they are the same. The combination of Ampanavos, Gallant, Chen, Meera, and Durand teaches a computer program product (Ampanavos, paragraph 7, line 3 “Second, we report on the results of an experiment and demonstrate that a machine learning-based system can successfully learn to generalize a consistent set of structural principles.” Examiner notes that one of ordinary skill in the art would understand the primary reference as teaching a computer program product comprising one or more computer readable storage media.“[I]n considering the disclosure of a reference, it is proper to take into account not only specific teachings of the reference but also the inferences which one skilled in the art would reasonably be expected to draw therefrom.” MPEP § 2144.01. In other words, one of ordinary skill in the art would know that a machine learning-based system is one or more computer readable storage media and program instructions stored on the one or more computer readable storage media.) Therefore, claims 15-17, and 19-20 are rejected for the same reasons as claims 1-3, and 6-7, respectively.
Claim 18 is a computer program product claim corresponding to the combination of computer method claims 4 and 5. Otherwise, it is the same. Therefore, claim 18 is rejected for the same reasons as clams 4 and 5.
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to BART RYLANDER whose telephone number is (571)272-8359. The examiner can normally be reached Monday - Thursday 8:00 to 5:30.
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, Miranda Huang can be reached at 571-270-7092. 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.
/Bart I Rylander/Examiner, Art Unit 2124