DETAILED ACTION
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Response to Arguments
Applicant’s arguments with respect to claims 1 - 20 have been considered but are moot because the new ground of rejection does not rely on any reference applied in the prior rejection of record for any teaching or matter specifically challenged in the argument.
Applicant argues that the prior art of record does not teach a predefined number of machine learning models; each machine learning model of the predefined number of machine learning models comprising a same type of machine learning model; wherein the set of vector representations for each word comprises the vector representation for each word output from each machine learning model; generating a perception score for each word based on the set of similarity scores for the set of vector representations for each word (Amendment, pages 8, 9).
Applicant's arguments filed 02/13/26 have been fully considered but they are not persuasive.
Applicant argues that the prior art of record does not teach generating a set of vector representations for each word of a plurality of words in social media data from a plurality of social network platforms in a predefined window of time by inputting each word into each machine learning model to output a vector representation for each word from each machine learning model (Amendment, pages 8, 9).
The examiner disagrees, since Arfa et al. disclose “A dictionary to map a plurality of unique n-grams to corresponding word vectors may be created using the trained word embedding model… An online social network may be able to train the word embedding model with a corpus of text generated by a large number of the online social network users…Vector Spaces and Embeddings”” (col.3, line 22 – col.4, line 67; col.36, lines 27 - 58).
Applicant argues that the prior art of record does not teach comparing each vector representation for each word to a given brand name to generate a similarity score for each vector representation and a set of similarity scores for each set of vector representations for each word (Amendment, pages 8, 9).
The examiner disagrees, since Arfa et al. disclose “A similarity metric of two vectors in the embedding space can be calculated. A similarity metric may be a cosine similarity, a Euclidean distance, a Jaccard similarity coefficient, or any suitable similarity metric. A similarity metric of two vectors may represent how the two corresponding n-grams are semantically similar to one another… The third-party user may send a request to the social-networking system 1860 to calculate similarity metrics for each pair of beer brands in the embedding space 1900”; col.3, line 22 – col.4, line 67; col.30, lines 33- 37).
Applicant argues that the amended claims include technical details that is not directed to any abstract idea (Amendment, page 8).
The examiner disagrees, and points out that the claims still recite an abstract idea.
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 - 20 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract without significantly more.
When considering subject matter eligibility under 35 USC 101, it must be determined whether the claim is directed to one of the four statutory categories of invention, i.e., process, machine, manufacture, or composition of matter.
Specifically, claims 1- 20 are directed to a method/system. They hereby fall under at least one of the four statutory classes of invention.
If the claim does not fall within one of the statutory categories, it must then be determined whether the claim is directed to a judicial exception (i.e., law of nature, natural phenomenon, and abstract idea).
Claims 1 - 20 recite steps of observation, evaluation, and judgement that can be practically performed by a human, either mentally or with the use of pen and paper.
The limitation of “generating a set of vector representations for each word of a plurality of words in social media data from a plurality of social network platforms in a predefined window of time by inputting each word into each machine learning model to output a vector representation for each word from each machine learning model; comparing each vector representation for each word to a given brand name to generate a similarity score for each vector representation and a set of similarity scores for each set of vector representations for each word; wherein the set of vector representations for each word comprises the vector representation for each word output from each machine learning model; generating a perception score for each word based on the set of similarity scores for the set of vector representations for each word” in claims 1- 20, is a process that, under its broadest reasonable interpretation, covers performance of the limitation in the mind, but for the recitation of generic computer components. That is, other than reciting “a memory that stores instructions; and one or more processors configured by the instructions to perform operations comprising a predefined number of machine learning models”, nothing in the claim element precludes the steps from practically being performed in a human mind. For example, but for the “a memory that stores instructions; and one or more processors configured by the instructions to perform operations comprising a predefined number of machine learning models” language, “generating a set of vector representations, generating a similarity score for each vector representation”, in the context of these claims encompasses the user mentally, or manually with the aid of pen and paper simply to generate a perception score for each word based on the generated similarity score for each vector representation in the set of vector representations for each word.
The mere nominal recitation of a memory that stores instructions; and one or more processors configured by the instructions to perform operations comprising a predefined number of machine learning models do not take the claim limitations out of the mental processes grouping.
If a claim limitation, under its broadest reasonable interpretation, covers mental processes but for the recitation of generic computer components, then it falls within the "Mental Processes" grouping of abstract ideas (concepts performed in the human mind including an observation, evaluation, judgement, and opinion). Accordingly, the claims recite an abstract idea.
This judicial exception is not integrated into a practical application. In particular, the claims recite the additional elements “inputting each word into each of a predefined number of machine learning models to output each of the set of vector representations”.
The limitation “inputting each word into each of a predefined number of machine learning models”, amount to data-gathering steps which is considered to be insignificant extra-solution activity, (See MPEP 2106.05(g)).
The limitation “output each of the set of vector representations”, represents an extra-solution activity because it is a mere nominal or tangential addition to the claim, a mere generic transmission and presentation of collected and analyzed data. (See MPEP 2106.05 (g)).
The claimed “a memory that stores instructions; and one or more processors configured by the instructions to perform operations comprising a predefined number of machine learning models” are recited at a high level of generality and are merely invoked as tool to perform an existing brand perception.
Accordingly, these additional elements do not integrate the abstract idea into a practical application because they do not impose any meaningful limits on practicing the abstract idea, thus fail to integrate the abstract idea into a practical application. See MPEP 2106.05(g).
The insignificant extra-solution activities identified above, which include the data-gathering (receiving, and extracting), and displaying steps, are recognized by the courts as well-understood, routine, and conventional activities when they are claimed in a merely generic manner (e.g., at a high level of generality) or as insignificant extra-solution activity (See MPEP 2106.05(d)(II) (i) Receiving or transmitting data over a network, e.g., using the Internet to gather data, buySAPE, Inc. v. Google, Inc., 765 F.3d 1350, 1355, 112 USPO2d 1093, 1096 (Fed. Cir. 2014) (computer receives and sends information over a network); (v) Presenting (displaying) offers and gathering statistics, OIP Techs., 788 F.3d at 1362-63, 115 USPO2d at 1092- 93). The claims are not patent eligible.
Claims 1 - 20 do not include additional elements that are sufficient to amount to significantly more than the judicial exception. As discussed above with respect to integration of the abstract idea into a practical application, the additional element of a memory that stores instructions; and one or more processors configured by the instructions to perform operations comprising a predefined number of machine learning models steps amounts to no more than mere instructions to apply the exception using a generic computer component. Mere instructions to apply an exception using a generic computer component cannot provide an inventive concept.
Even when considered in combination, these additional elements (a memory that stores instructions; and one or more processors configured by the instructions to perform operations comprising a predefined number of machine learning models) represent mere instruction to apply an exception and insignificant extra-solution activity, which do not provide an inventive concept.
Claims 1 - 20 as a whole, do not amount to significantly more than the abstract idea itself. This is because the claims do not affect an improvement to the functioning of a computer itself; and the claims do not move beyond a general link of the use of an abstract idea to a particular technological environment.
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 – 6, 8, 11 – 17, 19, and 20 are rejected under 35 U.S.C. 103 as being unpatentable over Arfa et al. (US Patent 10,558,759) in view of Gurbuxani et al.(US PAP 2021/0150541)
As per claims 1, 12, and 20, Arfa et al. teach a computer-implemented method, comprising:
generating a set of vector representations for each word of a plurality of words in social media data from a plurality of social network platforms in a predefined window of time by inputting each word into each machine learning model to output a vector representation for each word from each machine learning model (“A dictionary to map a plurality of unique n-grams to corresponding word vectors may be created using the trained word embedding model… An online social network may be able to train the word embedding model with a corpus of text generated by a large number of the online social network users…Vector Spaces and Embeddings”; col.3, line 22 – col.4, line 67; col.36, lines 27 - 58);
comparing each vector representation for each word to a given brand name to generate a similarity score for each vector representation and a set of similarity scores for each set of vector representations for each word (“A similarity metric of two vectors in the embedding space can be calculated. A similarity metric may be a cosine similarity, a Euclidean distance, a Jaccard similarity coefficient, or any suitable similarity metric. A similarity metric of two vectors may represent how the two corresponding n-grams are semantically similar to one another.”; col.3, line 22 – col.4, line 67).
However, Arfa et al. do not specifically teach a predefined number of machine learning models; each machine learning model of the predefined number of machine learning models comprising a same type of machine learning model; wherein the set of vector representations for each word comprises the vector representation for each word output from each machine learning model; generating a perception score for each word based on the set of similarity scores for the set of vector representations for each word.
Gurbuxani et al. disclose a positive word receives a positive score, a neutral word receives a score of 0 and a negative word receives a negative score. In some embodiments, all scores for each post are then added together, producing a final score used to determine the resulting sentiment. Positive score suggests a positive sentiment, negative and neutral scores—negative and neutral sentiments respectively…the vector is provided as input to a seventh ML model for processing. In some embodiments, the seventh ML model is trained on product and movie reviews. At step 1304c, the emoji features are provided as input to an eighth ML model for processing. In some embodiments, the seventh ML model is trained on Twitter messages. The seventh and eighth ML models may be used simultaneously to predict a sentiment category—positive, negative or neutral (paragraphs 304, 604, 605).
Therefore, it would have been obvious to one of ordinary skill in the art before the effective date of the filling claimed invention to use machine learning models for perception vectors as taught by Gurbuxani et al. in Arfa et al., because that would help predict sentiments of comments in social media posts by users (paragraph 604).
As per claims 2, 13, Arfa et al. in view of Gurbuxani et al. further disclose before generating the set of vector representations for each word, the method comprising: training each of the predefined number of machine learning models on the same training data to generate each of the predefined number of machine learning models configured to output a vector representation for a word (“The word embedding model may be a machine learning model (e.g., a neural network). An online social network may be able to train the word embedding model with a corpus of text generated by a large number of the online social network users… The seventh and eighth ML models may be used simultaneously to predict a sentiment category—positive, negative or neutral”; Arfa et al. col.1, lines 39 – 67; Gurbuxani et al; paragraphs 304, 604, 605).
As per claims 3, 14, Arfa et al. in view of Gurbuxani et al. further disclose the same type of machine learning model is a Word2Vec, FastText or DOeBERTs machine learning model (“The word embedding model may be trained using a word embedding training framework (e.g., Fasttext).”; Arfa et al., col.4, lines 5 – 12; Gurbuxani et al; paragraphs 304, 604, 605).
As per claims 4, 15, Arfa et al. in view of Gurbuxani et al. further disclose the similarity score is generated using cosine similarity to measure how important a word is with respect to the given brand name (“A similarity metric may be a cosine similarity, a Euclidean distance, a Jaccard similarity coefficient, or any suitable similarity metric. A similarity metric of two vectors may represent how the two corresponding n-grams are semantically similar to one another…A similarity metric of two vectors in the embedding space can be calculated. A similarity metric may be a cosine similarity, a Euclidean distance, a Jaccard similarity coefficient, or any suitable similarity metric. A similarity metric of two vectors may represent how the two corresponding n-grams are semantically similar to one another.”; Arfa et al., col.3, line 22 – col.4, line 67).
As per claims 5, 16, Arfa et al. in view of Gurbuxani et al. further disclose generating a perception score for each word comprises calculating an average of the set of similarity scores for set of vector representations for each word (“Brand Perception Overlap… The response message may include calculated similarity metrics corresponding to all the pairs of the word vectors. In particular embodiments, the response message may include instructions to display the calculated similarity metrics. The calculated similarity metrics may be color-coded where a color may represent any number within a pre-determined range. Although this disclosure describes identifying a similarity in public sentiments for each pair from a plurality of entities in a particular manner, this disclosure contemplates identifying a similarity in public sentiments for each pair from a plurality of entities in any suitable manner.”; Arfa et al., col.29, line 42 -col.30, line 67).
As per claims 6, 17, Arfa et al. in view of Gurbuxani et al. further disclose generating a perception score for each word comprises: for all vector representations generated by each of a given machine learning model of the predefined number of machine learning models, ranking the vector representations by similarity score and generating a ranked score for each vector representation based on the ranking; and generating the perception score for each word by calculating, for each word, an average of all ranked scores for the vector representations in the set of vector representations for each word (“The social-networking system 1860 may calculate an average vector 403 by taking a weighted average of word vectors 401 and 402, where an IDF score for the corresponding word is the weight applied to a word vector…The social-networking system 1860 may condense the d-dimensional word vectors in the table into a two-dimensional word vectors by performing a t-distributed Stochastic Neighbor Embedding (t-SNE) dimensionality reduction on the word vectors in the first table. The social-networking system 1860 may determine a Term Frequency-Inverse Document Frequency (TF-IDF) ranking of the n-grams in the first table.” Arfa et al., col.11, lines 5 – 42, col.13, lines 23 – 67).
As per claim 8, Arfa et al. in view of Gurbuxani et al. further disclose before generating a set of vector representations for each word of a plurality of words in a set of social media data, the method comprises: extracting, from a plurality of social network platforms, social media data related to a given brand; and performing data cleaning to generate the set of social media data (Arfa et al., col.3, line 22 – col.4, line 67).
As per claims 11, 19, Arfa et al. in view of Gurbuxani et al. further disclose identifying top concepts based on the perception score for each word; and comparing the top concepts to previously determined concepts in previous time periods (“the social-networking system 1860 may receive a request to generate k words that each approximates a representation of a relationship between two concepts from a computing device. The online social network may have a large corpus of text collected from content objects generated by users. Because a number of users generating the content objects may be large and the users may be well distributed in terms of demographics, the corpus of text may represent contemporary public sentiments. A third-party user associated with the computing device may want to understand a relationship between two concepts.”; Arfa et al., col.9, line 40 -col.10, line 67).
Conclusion
Applicant's amendment necessitated the new ground(s) of rejection presented in this Office action. Accordingly, THIS ACTION IS MADE FINAL. See MPEP § 706.07(a). Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any nonprovisional extension fee (37 CFR 1.17(a)) pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to LEONARD SAINT-CYR whose telephone number is (571)272-4247. The examiner can normally be reached Monday- Friday.
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, Richemond Dorvil can be reached at (571)272-7602. 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.
/LEONARD SAINT-CYR/Primary Examiner, Art Unit 2658