Detailed Action
This action is written in response to 6/25/24. The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Subject Matter Eligibility
In determining whether the claims are subject matter eligible, the examiner has considered and applied guidance from MPEP § 2106. The examiner finds that the independent claims are directed to the practical application of providing a user interface for interactively controlling a specific machine learning process. Specifically, the system generates predictions using an ML model, then allows the user to amend these predictions, and the model is subsequently retrained using the user’s feedback. The ML model can then generate new predictions based upon the updated model. Furthermore, the combination of steps performed in the recited method cannot be practically performed as a mental process.
Double Patenting
The nonstatutory double patenting rejection is based on a judicially created doctrine grounded in public policy (a policy reflected in the statute) so as to prevent the unjustified or improper timewise extension of the “right to exclude” granted by a patent and to prevent possible harassment by multiple assignees. A nonstatutory double patenting rejection is appropriate where the conflicting claims are not identical, but at least one examined application claim is not patentably distinct from the reference claim(s) because the examined application claim is either anticipated by, or would have been obvious over, the reference claim(s). See, eg, In re Berg, 140 F.3d 1428, 46 USPQ2d 1226 (Fed. Cir. 1998).
A timely filed terminal disclaimer in compliance with 37 CFR 1.321(c) or 1.321(d) may be used to overcome an actual or provisional rejection based on nonstatutory double patenting provided the reference application or patent either is shown to be commonly owned with the examined application, or claims an invention made as a result of activities undertaken within the scope of a joint research agreement. See MPEP § 717.02 for applications subject to examination under the first inventor to file provisions of the AIA as explained in MPEP § 2159. A terminal disclaimer must be signed in compliance with 37 CFR 1.321(b).
The filing of a terminal disclaimer by itself is not a complete reply to a nonstatutory double patenting (NSDP) rejection. A complete reply requires that the terminal disclaimer be accompanied by a reply requesting reconsideration of the prior Office action. Even where the NSDP rejection is provisional the reply must be complete. See MPEP § 804, subsection I.B.1. For a reply to a non-final Office action, see 37 CFR 1.111(a). For a reply to final Office action, see 37 CFR 1.113(c). A request for reconsideration while not provided for in 37 CFR 1.113(c) may be filed after final for consideration. See MPEP §§ 706.07(e) and 714.13.
The USPTO Internet website contains terminal disclaimer forms which may be used. Please visit www.uspto.gov/patent/patents-forms. The actual filing date of the application in which the form is filed determines what form (e.g., PTO/SB/25, PTO/SB/26, PTO/AIA /25, or PTO/AIA /26) should be used. A web-based eTerminal Disclaimer may be filled out completely online using web-screens. An eTerminal Disclaimer that meets all requirements is auto-processed and approved immediately upon submission. For more information about eTerminal Disclaimers, refer to www.uspto.gov/patents/apply/applying-online/eterminal-disclaimer.
Claims 1-18 are rejected on the ground of nonstatutory double patenting as being unpatentable over the corresponding claims of U.S. Patent No. 12,242,954. Although the claims at issue are not identical, they are not patentably distinct from each other for the reasons outlined in the table below.
This application – 18/753483
Issued Patent – US 12,242,954 B2
1. A system comprising:
1. A system comprising:
a processor; and
a processor; and
a memory storing instructions that, when executed by the processor, configure the system to:
a memory storing instructions that, when executed by the processor, configure the system to:
pre-process, by a machine learning module, data;
convert a plurality of descriptions of items into a plurality of word vectors, each word vector having a plurality of dimensions;
project, onto a two-dimensional plane of a user interface, the plurality of word vectors;
select, by the machine learning module, a trained machine learning model;
train a neural network on the plurality of word vectors and a plurality of sets of two-dimensional coordinates, each set of two-dimensional coordinates associated with a respective word vector;
predict, by the machine learning module, a result based on the trained machine learning model;
predict a result based on the trained neural network;
output, by the machine learn module, to a user interface, a prediction for a user;
output, to the user interface, a prediction comprising cluster groupings of the plurality of sets of two dimensional coordinates;
amend, via the user interface, the prediction, by the user, to provide an amended prediction;
amend, via the user interface and in response to user input, the prediction, to provide an amended prediction, wherein the amending comprises moving on a screen displaying the user interface a subset of the plurality of sets of two-dimensional coordinates from one of the cluster grouping to another one of the cluster groupings or to a new cluster grouping;
retrain, by the machine learning module, the trained machine learning model based on data associated with the amended prediction, thereby providing a retrained machine learning model; and
retrain the trained neural network based on the amended prediction, thereby providing a retrained neural network; and
predict, by the machine learning module, a new result based on: (i) the data associated with the amended prediction; and (ii) the re-trained machine learning model.
predict a new result based on the retrained neural network.
As illustrated in the table above, every limitation in this application has a corresponding equivalent or more specific limitation in the ‘954 patent. Thus, the ‘954 patent anticipates this claim.
Independent claims 7 and 13 recite an analogous method and an analogous non-transitory computer-readable storage medium, respectively.
The correspondence in dependent claims is illustrated in the table below.
This application – 18/753483
Issued Patent – US 12,242,954 B2
2. The system of claim 1, wherein the user interface is a graphical user interface.
2. The system of claim 1, wherein the user interface is a graphical user interface.
3. The system of claim 2, wherein the results are output to a device; and the user amends the prediction by moving one or more objects on a screen of the device.
3. The system of claim 2, wherein the results are output to a device.
4. The system of claim 1, wherein the user amends the prediction by amending a data file associated with the prediction.
4. The system of claim 1, wherein the amending the prediction comprises amending, in response to user input, a data file associated with the prediction.
5. The system of claim 1, wherein the machine learn model is selected from the group consisting of K-Means Clustering, Mean-Shift Clustering, Density-Based Spatial Clustering of Applications with Noise (DBSCAN), Expectation–Maximization (EM) Clustering using Gaussian Mixture Models (GMM), Agglomerative Hierarchical Clustering and any combination thereof.
The ’954 does not recite this feature in the claimset.
However, this limitation is taught by Runkler, see p. 993 (excerpt reproduced infra in §103 rejections).
At the time of filing, it would have been obvious to a skilled machine learning engineer to modify the claims of the ‘954 patent to include the k-means clustering algorithm (as taught by Runkler) because this is the most popular algorithm for clustering.
6. The system of claim 1, wherein the instructions configure the system to:
convert, by the machine learning module, a plurality of descriptions of items into a plurality of word vectors, each word vector having a plurality of dimensions;
project, by the machine learning module onto a two-dimensional plane of the user interface, the plurality of word vectors;
train, by the machine learning module, a neural network within the machine learning module, on the plurality of word vectors and a plurality of sets of two-dimensional coordinates, each set of two-dimensional coordinates associated with a respective word vector;
amend, by the user via the user interface, a subset of the plurality of sets of two-dimensional coordinates, to provide a plurality of amended sets of two-dimensional coordinates; andretrain, by the machine learning module, the neural network, on the plurality of amended sets of two-dimensional coordinates.
1. A system comprising:
a processor; and
a memory storing instructions that, when executed by the processor, configure the system to:
convert a plurality of descriptions of items into a plurality of word vectors, each word vector having a plurality of dimensions;
project, onto a two-dimensional plane of a user interface, the plurality of word vectors;
train a neural network on the plurality of word vectors and a plurality of sets of two-dimensional coordinates, each set of two-dimensional coordinates associated with a respective word vector; ….
amend, via the user interface and in response to user input, the prediction, to provide an amended prediction, wherein the amending comprises moving on a screen displaying the user interface a subset of the plurality of sets of two-dimensional coordinates from one of the cluster grouping to another one of the cluster groupings or to a new cluster grouping;
retrain the trained neural network based on the amended prediction, thereby providing a retrained neural network; and
predict a new result based on the retrained neural network.
Regarding claims 2-4 and 6, as illustrated in the table above, every claim limitation in this application has a corresponding equivalent or more specific limitation in the ‘954 patent. Thus, the ‘954 patent anticipates these claim.
Regarding dependent claim 5, As illustrated in the table above, most limitations in this application has a corresponding equivalent or more specific limitation in the ‘954 patent; the Runkler reference
discloses the remaining limitations. (See obviousness statement above.)
Dependent claims 8-12 and 14-18 are analogous to dependent claims 2-6.
Claim Objections
The Examiner objects to claim 1, which recites “output, by the machine learn module”. This should be amended to read “output, by the machine learning module …”. Independent claim 13 has the same error. Appropriate correction is required.
Claim Rejections - 35 USC § 103
The following is a quotation of 35 U.S.C. 103(a) which forms the basis for all obviousness rejections set forth in this Office action:
(a) A patent may not be obtained though the invention is not identically disclosed or described as set forth in section 102 of this title, if the differences between the subject matter sought to be patented and the prior art are such that the subject matter as a whole would have been obvious at the time the invention was made to a person having ordinary skill in the art to which said subject matter pertains. Patentability shall not be negatived by the manner in which the invention was made.
The following references are relied upon in the rejections below:
Anthru (US 2012/0311637 A1. Cited by Applicant on IDS dated 3/20/25.)
PC Review (PC Review discussion forum, “On scatter chart, how can data points be moved with curser?”, posts dated 2/8/10, available at https://www.pcreview.co.uk/threads/on-scatter-chart-how-can-data-points-be-moved-with-curser.3965453/, accessed 9/11/23. 1 page. Cited by Applicant on IDS dated 3/20/25.)
Runkler (Runkler TA. Generation of linguistic membership functions from word vectors. In 2016 IEEE International Conference on Fuzzy Systems (FUZZ-IEEE) 2016 Jul 24 (pp. 993-999). IEEE. Cited by Applicant on IDS dated 3/20/25.)
Claims 1-18 are rejected under 35 U.S.C. 103 as being unpatentable over Anthru, Runkler and PC Review.
Regarding claims 1, 7, and 13, Anthru discloses a system comprising:
a processor; and
[0020] processor.
a memory storing instructions that, when executed by the processor, configure the system to: …
[0020] memory
select, by the machine learning module, a trained machine learning model;
[0007] “A cluster processor determines the relationship between each identified program based on the at least one program description attribute and generates at least one cluster including at least one of the plurality of identified programs based on the determined relationship.”
predict, by the machine learning module, a result based on the trained machine learning model;
[0024] “the present system advantageously generates clusters of related pieces of content from a set of available content by considering all dimensions to which the content belongs.”
output, by the machine learn module, to a user interface, a prediction for a user; …
[0012] “FIG. 3 is an exemplary display image of a clustered program guide in accordance with an aspect of an embodiment”.
[0024] “the present system advantageously generates clusters of related pieces of content from a set of available content by considering all dimensions to which the content belongs.”
retrain, by the machine learning module, the trained machine learning model based on data associated with the amended prediction, thereby providing a retrained machine learning model; and
[0092] “In a further embodiment, the UI display processor 840 may format the attribute data to enable selective modification thereof in response to user action.… In response to receiving a modification signal from a user modifying at least one attribute of the Complete Attribute Set, the cluster processor 825 automatically executes the clustering algorithm to re-cluster and determine the relatedness of the Sample Set of Available Programs based on the modified Complete Attribute Set and provides the modified re-clustered data to the UI display processor 840 for display thereof.” (Emphasis added.)
predict, by the machine learning module, a new result based on: (i) the data associated with the amended prediction; and (ii) the re-trained machine learning model.
[0092] “In response to receiving a modification signal from a user modifying at least one attribute of the Complete Attribute Set, the cluster processor 825 automatically executes the clustering algorithm to re-cluster and determine the relatedness of the Sample Set of Available Programs based on the modified Complete Attribute Set and provides the modified re-clustered data to the UI display processor 840 for display thereof.” (Emphasis added.)
Runkler discloses the following limitations which Anthru does not disclose:
pre-process, by a machine learning module, data;
P. 994, sec. II Word vectors “In the field of natural language processing there are currently many activities to generate so–called word vector data sets from huge text corpora. The idea is to consider the 𝑛 different words of a text corpus and count how often each pair of these words occurs together in the same section or column of the corpus. The resulting symmetric 𝑛 × 𝑛 co–occurence matrix is factorized, which yields a 𝑝–dimensional vector representation”. (Emphasis added.)
P. 995, second col. “In Fig. 2 we used two-dimensional PCA to project the 50–dimensional word vectors of 20 countries to the two-dimensional plane, so that similar words were placed close to each other”. (Emphasis added.)
At the time of filing, it would have been obvious to a person of ordinary skill to combine the techniques disclosed by Runkler for word vector representation of text data with the Anthru system because the former make unstructured text data usable by machine learning models (eg neural networks) for common tasks including classification and clustering. Both disclosures pertain to machine learning.
PC Review discloses the following limitation which neither Anthru nor Runkler discloses:
amend, via the user interface, the prediction, by the user, to provide an amended prediction;
P. 1 “EXCEL 2003 allows data point a on a scatter chart to be dragged from one location to another. I use this feature a lot to fit an equation with dummy data points to actual data points. I am unable to do this in EXCEL 2007. I have asked others with this new system and they were not able find a way to do it either. Has this function been removed? How can this feature be activated.”(cont.) “According to MS documentation:-
Direct manipulation of data points: In the past, you could select a data point in a chart and move it — which also changed the underlying data. This rarely used action no longer works."
At the time of filing, it would have been obvious to a person of ordinary skill to apply the technique disclosed by PC Review for moving data points via a GUI to the combined system of Anthru/Runkler because this gives the user control over the data points being used to generate the model, and therefore control over the model itself. Although PC Review does not disclose moving a data point “from one of the cluster groupings to another one of the cluster groupings”, this could be intuitively achieved if the data points were already clustered, as in Anthru (see excerpt from fig. 7 below).
PNG
media_image1.png
428
522
media_image1.png
Greyscale
Regarding independent claim 13, Antrhu further discloses a non-transitory computer-readable storage medium.
[0096] “Additionally, the methods may be implemented by instructions being performed by a processor, and such instructions may be stored on a processor or computer-read able media such as, for example, an integrated circuit, a Software carrier or other storage device such as, for example, a hard disk, a compact diskette, a random access memory (“RAM), a read-only memory (“ROM) or any other magnetic, optical, or solid state media.”
Regarding claims 2, 8 and 14, Anthru discloses the further limitation wherein the user interface is a graphical user interface.
See excerpt from fig. 7, reproduced above, showing a screenshot of a GUI.
Regarding claims 3, 9 and 15, Anthru discloses the further limitation wherein the results are output to a device.
Fig. 2 and [0049]: primary and secondary display device.
PC Review discloses the following limitation which neither Anthru nor Runkler discloses:
the user amends the prediction by moving one or more objects on a screen of the device.
P. 1 “EXCEL 2003 allows data point a on a scatter chart to be dragged from one location to another. I use this feature a lot to fit an equation with dummy data points to actual data points. I am unable to do this in EXCEL 2007. I have asked others with this new system and they were not able find a way to do it either. Has this function been removed? How can this feature be activated.”(cont.) “According to MS documentation:-
Direct manipulation of data points: In the past, you could select a data point in a chart and move it — which also changed the underlying data. This rarely used action no longer works."
Regarding claims 4, 10 and 16, PC Review discloses the further limitation wherein the user amends the prediction by amending a data file associated with the prediction.
P. 1 “EXCEL 2003 allows data point a on a scatter chart to be dragged from one location to another. I use this feature a lot to fit an equation with dummy data points to actual data points. I am unable to do this in EXCEL 2007. I have asked others with this new system and they were not able find a way to do it either. Has this function been removed? How can this feature be activated.”(cont.) “According to MS documentation:-
Direct manipulation of data points: In the past, you could select a data point in a chart and move it — which also changed the underlying data. This rarely used action no longer works."
Regarding claims 5, 11 and 17, Runkler discloses the further limitation wherein the machine learn model is selected from the group consisting of K-Means Clustering, Mean-Shift Clustering, Density-Based Spatial Clustering of Applications with Noise (DBSCAN), Expectation–Maximization (EM) Clustering using Gaussian Mixture Models (GMM), Agglomerative Hierarchical Clustering and any combination thereof.
P. 993, second col., “Some of the most popular methods for unsupervised generation of membership functions are based on fuzzy clustering [5]. For example, Pal et al. [6] use fuzzy c–means clustering for generating membership functions and rules, and Runkler et al. [7] use fuzzy c–elliptotype clustering for generating Takagi–Sugeno type fuzzy systems [8].” (Emphasis added.)
Regarding claims 6, 12 and 18, Runkler discloses the following limitations which Anthru does not disclose wherein the instructions configure the system to: …
convert, by the machine learning module, a plurality of descriptions of items into a plurality of word vectors, each word vector having a plurality of dimensions;
P. 994, sec. II Word vectors “In the field of natural language processing there are currently many activities to generate so–called word vector data sets from huge text corpora. The idea is to consider the 𝑛 different words of a text corpus and count how often each pair of these words occurs together in the same section or column of the corpus. The resulting symmetric 𝑛 × 𝑛 co–occurence matrix is factorized, which yields a 𝑝–dimensional vector representation”. (Emphasis added.)
P. 995, second col. “In Fig. 2 we used two-dimensional PCA to project the 50–dimensional word vectors of 20 countries to the two-dimensional plane, so that similar words were placed close to each other”. (Emphasis added.)
project, by the machine learning module onto a two-dimensional plane of the user interface, the plurality of word vectors;
Id.
train, by the machine learning module, a neural network within the machine learning module, on the plurality of word vectors and a plurality of sets of two-dimensional coordinates, each set of two-dimensional coordinates associated with a respective word vector;
P. 993, “Many pure supervised approaches to learning membership functions use neural networks. Neural networks are used to tune membership function parameters [11] or to learn membership functions [12]. Specific neural network architectures have been proposed that represent entire fuzzy systems, for example NEFCON/NEFCLASS [13] or ANFIS [14].”
PC Review discloses the following limitation which neither Anthru nor Runkler discloses:
amend, by the user via the user interface, a subset of the plurality of sets of two-dimensional coordinates, to provide a plurality of amended sets of two-dimensional coordinates; and
P. 1 “EXCEL 2003 allows data point a on a scatter chart to be dragged from one location to another. I use this feature a lot to fit an equation with dummy data points to actual data points. I am unable to do this in EXCEL 2007. I have asked others with this new system and they were not able find a way to do it either. Has this function been removed? How can this feature be activated.”(cont.) “According to MS documentation:-
Direct manipulation of data points: In the past, you could select a data point in a chart and move it — which also changed the underlying data. This rarely used action no longer works."
Anthru discloses the following further limitation:
retrain, by the machine learning module, the neural network, on the plurality of amended sets of two-dimensional coordinates.
[0092] “In a further embodiment, the UI display processor 840 may format the attribute data to enable selective modification thereof in response to user action.… In response to receiving a modification signal from a user modifying at least one attribute of the Complete Attribute Set, the cluster processor 825 automatically executes the clustering algorithm to re-cluster and determine the relatedness of the Sample Set of Available Programs based on the modified Complete Attribute Set and provides the modified re-clustered data to the UI display processor 840 for display thereof.” (Emphasis added.)
Additional Relevant Prior Art
The following references were identified by the Examiner as being relevant to the disclosed invention, but are not relied upon in any rejection:
Du discloses an application of neural networks to a clustering task. (Du KL. Clustering: A neural network approach. Neural networks. 2010 Jan 1;23(1):89- 107. Cited by Applicant on IDS dated 3/20/25.)
Kumar discloses an in-depth discussion of the k-means clustering algorithm. See p. 75 et seq., sec. 3.4.1.3. (Bhattacharyya, Dhruba Kumar, and Jugal Kumar Kalita. Network anomaly detection: A machine learning perspective. Crc Press, 2013.)
Conclusion
Information regarding the status of an application may be found at the USPTO Patent Center at https://patentcenter.uspto.gov. Any inquiry concerning this communication should be directed to Vincent Gonzales at (571) 270-3837. The examiner can normally be reached Monday-Friday 7 am to 4 pm MT. If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Miranda Huang, can be reached at (571) 270-7092.
/Vincent Gonzales/Primary Examiner, Art Unit 2124
