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 .
Claims 81-82, 86-89, 93-96, 100 and 215-217 are pending.
Response to Arguments
Applicant’s arguments filed, with respect to the rejection of Claims 81-82, 86-89, 93-96 and 100 under 35 USC 101, have been fully considered and are persuasive. The independent claims (Claims 81, 88 and 95) have been amended to include additional features and limitations to claim a patent eligible subject matter that is significantly more than an abstract idea. Therefore, the rejection of Claims 81-82, 86-89, 93-96 and 100 under 35 USC 101 has been withdrawn.
Applicant’s arguments filed, with respect to the amended independent claims (Claims 81, 88 and 95) and the previously cited prior art, have been fully considered and are persuasive. Therefore, the rejection of Claims 81-82, 86-89, 93-96 and 100 under 35 USC 103 has been withdrawn. However, upon further consideration, a new ground of rejection is made in view of Shalev et al. (US Pub. 2023/0034085 A1) and Begun et al. (US Pub. 2021/0081613 A1).
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 81, 86-88, 93-95 and 100 are rejected under 35 U.S.C. 103 as being unpatentable over Subramaniam et al. (US Patent 11,720,964 B2) in view of Shalev et al. (US Pub. 2023/0034085 A1), and in further view of Begun et al. (US Pub. 2021/0081613 A1).
Regarding Claims 81, 88 and 95, Subramaniam teaches a method of automatically populating a form (see Fig.4, Fig.5 and Col.15, Line 8-10), comprising:
selecting the form to populate based on an aspect of an interaction, the form including a plurality of input fields (see Fig.5 ($502,5504,S506) and Col.15, Line 11-26, obtaining or selecting a trading ticket with data fields to automatically populate based on user input information);
generating a record of an interaction including content, the content being attributed to a participant in the interaction (see Fig.5 (S504) and Col.15, Line 19-23, obtaining the user input data to interpret and extract key data for processing);
and populating the plurality of input fields with data associated with a corresponding input field using a form filling model (see Fig.4 (406,418), Fig.5 (506), Col.14, Line 17-20 and Col.15, Line 31-34, the natural language order fill module is configured as an LSTM neural network model),
wherein the form filling model is trained using a dataset including filled out forms having answers in populated input fields (see Fig.4 (412,422) and Col.14, Line 49 – Col.15, Line 7, the form filling model is trained with training tickets stored in the database and learning through time), the training of the form filling model including extracting features from the dataset (see Fig.4 (422) and Col.14, Line 49-55, extracting key trade details from trading tickets),
and wherein populating the plurality of input fields using the form filling model comprises extracting features from content parts (see Fig.5 (S504) and Col.15, Line 19-23, extracting key trade details), and populating an input field of the selected form with one or more content parts (see Fig.4 (406), Fig.5 (S506) and Col.14, Line 44-48, mapping key trade details to form fields).
Subramaniam fails to teach generating a transcript of the interaction, splitting the content of the transcript into content parts, clustering the extracted features into a plurality of clusters, wherein one or more clusters of the plurality of clusters corresponds to each question of the filled out forms and the extracted features comprising word embeddings for the questions and respective answers forms, and using a similarity measure to compare the extracted features.
Shalev, however, teaches generating a transcript of a conversation that includes word embeddings to indicate a level semantic similarity between the words of the transcript (see Fig.2 (222,225,228), paragraph [0023] and paragraph [0041]), wherein the conversation duration can be divided and participant embeddings can be generated for the transcript (see paragraph [0031] and paragraphs [0071-0072]).
Begun teaches using a neural network model to label key features of a document, clustering the labels or embeddings, and using similarity metrics and word embeddings to map semantic roles and patterns of the key features in the document (see paragraphs [0053-0059]).
It would have been obvious for one skilled in the art, before the effective filing date of the application, to include to Subramaniam’s method the steps for generating a transcript of the interaction, generating word embeddings for the transcript and splitting the content of the transcript into content parts. The motivation would be to identify portions of the transcript corresponding to the participants of the interaction, and to identify the similarity between the words of the interaction.
It is further obvious to include to Subramaniam’s method the steps for clustering the extracted features into a plurality of clusters, wherein one or more clusters of the plurality of clusters corresponds to each question of the filled out forms and the extracted features comprising word embeddings for the questions and respective answers forms, and using a similarity measure to compare the extracted features. The motivation would be to use a machine learning model to group words with similar meanings, semantic roles or context, and to map the answers to the question or input fields of the forms.
Regarding Claims 86, 93 and 100, Subramaniam further teaches wherein the content comprises non-word symbols and language characteristics (see Fig.5 (s502) and Col.8, Line 10-22, abbreviations, sentence syntaxes and special characters as input data).
Regarding Claims 87 and 94, Subramaniam further teaches wherein the form filling model is a natural language processing (NLP) model (see Fig.4 (406) and Col.14, Line 64 - Col.15, Line 4, natural language order fill module is configured as an LSTM neural network model that processes natural language input data).
Claims 82, 89 and 96 are rejected under 35 U.S.C. 103 as being unpatentable over Subramaniam et al. (US Patent 11,720,964 B2) in view of Shalev et al. (US Pub. 2023/0034085 A1), and in further view of Begun et al. (US Pub. 2021/0081613 A1), and in further view of Le et al. (US Patent 10,431,205 B2).
Regarding Claims 82, 89 and 96, Subramaniam, Shalev and Begun teach the method of Claim 81 but fail to teach wherein the form filling model comprises a plurality of models with at least one model for each question associated with an input field of the form.
Le, however, teaches utilizing multiple language models to provide responses to questions based on context or domain-specific information (see Fig. (30,42) and Col.5, Line 33-44).
It would have been obvious for one skilled in the art, before the effective filing date of the application, to include to the method of Claim 81 the step for using a plurality of models with at least one model for each question associated with an input field of the form. The motivation would be to provide answers to the question data fields associated with different domains, categories, topics or industries.
Claims 215-217 are rejected under 35 U.S.C. 103 as being unpatentable over Subramaniam et al. (US Patent 11,720,964 B2) in view of Shalev et al. (US Pub. 2023/0034085 A1), and in further view of Begun et al. (US Pub. 2021/0081613 A1), and in further view of Liu et al. (US Pub. 2021/0125085 A1).
Regarding Claims 215, 216 and 217, Subramaniam, Shalev and Begun teach the method of Claim 81 but fail to teach wherein the word embeddings each represent data in pre-defined length regardless of a word or number of words in a phrase.
Liu, however, teaches generating word embeddings of predefined lengths for outputting or generating inferred embeddings for incomplete words (see Fig.12 and paragraphs [0063-0064]).
It would have been obvious for one skilled in the art, before the effective filing date of the application, to include to the method of Claim 81 the step for generating word embeddings each represent data in pre-defined length regardless of a word or number of words in a phrase. The motivation would be to use a machine learning model to predict candidate words for incomplete words in the interaction transcript.
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 VU B HANG whose telephone number is (571)272-0582.
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, Hai Phan, can be reached at (571)272-6338. 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.
/VU B HANG/Primary Examiner, Art Unit 2654
