DETAILED ACTION
This communication is in response to the application filed 10/11/23 in which claims 1-20 were presented for examination.
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 .
Information Disclosure Statement
The information disclosure statements (IDS) submitted on 4/19/24 and 6/21/24 are in compliance with the provisions of 37 CFR 1.97. Accordingly, the information disclosure statements are being considered by the examiner.
Claim Rejections - 35 USC § 112
The following is a quotation of 35 U.S.C. 112(b):
(b) CONCLUSION.—The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the inventor or a joint inventor regards as the invention.
The following is a quotation of 35 U.S.C. 112 (pre-AIA ), second paragraph:
The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the applicant regards as his invention.
Claim 10 is rejected under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), second paragraph, as being indefinite for failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor (or for applications subject to pre-AIA 35 U.S.C. 112, the applicant), regards as the invention.
Claim 10 recites “the first user can view the electronic signature in near real-time relative to when the field assignee input the electronic signature at the field.” The term “near real-time” is indefinite at least because its limits cannot be determined objectively.
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 idea without significantly more. Specifically, the claims describe generating a second portion of a natural language sentence on a form based on a first portion of the sentence. The context of such a document drafting exercise arises regularly in the legal field and is performed mentally. Such a completion of a first portion of a natural language text with a second portion may be performed by evaluation, opinion, and judgment and, therefore, falls under the Mental Processes grouping of abstract ideas. As indicated in the more detailed analysis below, the additional elements of the claims fail to integrate the exception into a practical application or provide an inventive concept.
Claim 1
A computerized system comprising: [Step 1: YES, claim is directed to a system which falls under a statutory category]
one or more processors; and [Step 2A Prong 2/Step 2B: NO. Recitation of generic computer components does not integrate the judicial exception into a practical application]
computer storage memory having computer-executable instructions stored thereon which, when executed by the one or more processors, implement a method comprising: [Step 2A Prong 2/Step 2B: NO. Recitation of generic computer components does not integrate the judicial exception into a practical application]
receiving, over a computer network, a first request to generate one or more elements of a document of a web application, the document being native to the web application, the document requiring at least one electronic signature by one or more entities, the first request being issued at a user device associated with a first user; and [Step 2A Prong 2: NO. Receiving information/data is considered insignificant extra-solution activity. Step 2B: NO. Under 2B this insignificant extra solution activity is well understood routine and conventional activity. See “Receiving or transmitting data over a network, e.g., using the Internet to gather data, Symantec, 838 F.3d at 1321, 120 USPQ2d at 1362.”]
based at least in part on at least one of: the first request or user input of one or more natural language characters at the document, automatically causing generation, via a machine learning model, of at least one of:
(1) one or more strings at the document, or [Step 2A Prong 1: YES. Mental process capable of being performed by evaluation, opinion, and judgment. Step 2A Prong 2/Step 2B: NO. Recitation of generic computer components (e.g., automatically, via a machine learning model] is mere instruction to apply the exception.]
(2) a field, the field being a data object representing a predetermined category for which a second user is to input data within the field according to the predetermined category [Step 2A Prong 1: YES. Mental process capable of being performed by evaluation, opinion, and judgment. Step 2A Prong 2/Step 2B: NO. Recitation of generic computer components (e.g., automatically, via a machine learning model] is mere instruction to apply the exception.].
Claim 2
The computerized system of claim 1, wherein the one or more strings include at least one of:
a natural language sequence that represents a completed portion of the one or more natural language characters of the user input, [Step 2A Prong 1: YES. Mental process capable of being performed by evaluation, opinion, and judgment.]
a template representing pre-formatted natural language content that was generated at a prior time or session relative to a time or session associated with the receiving of the first request, or [Step 2A Prong 1: YES. Mental process capable of being performed by evaluation, opinion, and judgment.]
a name of an assignee of the field, wherein the assignee includes the second user [Step 2A Prong 1: YES. Mental process capable of being performed by evaluation, opinion, and judgment.].
Claim 3
The computerized system of claim 1, wherein the generation of at least one of: the one or more strings or the field is further based at least in part on a past history of computer inputs by the first user that input the one or more natural language characters [Step 2A Prong 1: YES. Mental process capable of being performed by evaluation, opinion, and judgment; Step 2A Prong 2/Step 2B: NO. Recitation of generic computer components (e.g., computer) is mere instruction to apply exception.].
Claim 4
The computerized system of claim 1, wherein the method of the computerized system further comprising:
causing presentation, at the document and at the user device, of a message that informs the first user that the at least one of the one or more strings or the field is a suggestion to input next to a partial string representing the one or more natural language characters [Step 2A Prong 2: Data outputting is insignificant extra-solution activity. The type or source of data does not cause the data outputting to integrate the exception into a practical application. Step 2B: NO. Under 2B this insignificant extra solution activity is well understood routine and conventional activity. See “Receiving or transmitting data over a network, e.g., using the Internet to gather data, Symantec, 838 F.3d at 1321, 120 USPQ2d at 1362.”].
Claim 5
The computerized system of claim 1, wherein the method of the computerized system further comprising;
at least partially in response to receiving the user input of the one or more natural language characters at the document, encoding the one or more natural language characters into one or more word embedding feature vectors that represent positional information or context of each word in the user input; and [Step 2A Prong 1: YES. Mental process capable of being performed by evaluation, opinion, and judgment.]
in response to the encoding, generating an attention vector for each word of the user input by determining how relevant, via weighting, each word in the user input is relative to at least one other word in the user input and the one of the one or more strings and the field, and wherein the generating of at least one of the one or more strings or the field is based on the generation of an attention vector for each word [Step 2A Prong 1: YES. Mental process capable of being performed by evaluation, opinion, and judgment.].
Claim 6
The computerized system of claim 1, wherein the method of the computerized system further comprising: training or fine-tuning the machine learning model by learning string pairs or string-field pairs, each string-field pair indicates at field that is predicted to be placed next to a certain string, each string pair indicates a first string that is predicted to be placed next to a second string [Step 2A Prong 1: YES. Mental process capable of being performed by evaluation, opinion, and judgment. Step 2A Prong 2/Step 2B: Recitation of generic computer components (e.g., computerized system, machine learning model) is mere instruction to apply exception.].
Claim 7
The computerized system of claim 1, wherein the method further comprises:
as part of the automatic generation of at least one of the one or more strings or the field, automatically moving at least one of: the one or more natural language characters, the one or more strings, or the field from a first line to a second line in the document based at least in part on a screen size of the user device [Step 2A Prong 1: YES. Mental process capable of being performed by evaluation, opinion, and judgment. Step 2A Prong 2/Step 2B: Recitation of generic computer components or language (e.g., automatic, screen, device) is mere instruction to apply exception.].
Claim 8
The computerized system of claim 1, wherein the method of the computerized system further comprises:
subsequent to the automatic generation of at least one of the one or more strings or the field, assigning the field to the second user based on second computer user input from the first user, the assigning being indicative of authorizing only the second user, and not the first user, to populate the field; and [Step 2A Prong 1: YES. Mental process capable of being performed by evaluation, opinion, and judgment.]
in response to the assigning, automatically causing a transmission, over the computer network, of an indication to a device associated with the second user, wherein the second user is able to populate the field, at the document, based on the assigning of the field and the transmission of the indication [Step 2A Prong 2: Data outputting is insignificant extra-solution activity. The type or source of data does not integrate the judicial exception into a practical application. Step 2B: NO. Under 2B this insignificant extra solution activity is well understood routine and conventional activity. See “Receiving or transmitting data over a network, e.g., using the Internet to gather data, Symantec, 838 F.3d at 1321, 120 USPQ2d at 1362.”].
Claim 9
The computerized system of claim 1, wherein the field is one of: a signature field, a company name field, a personal name field, an email address field, a job title field, or a residential or business address field [Step 2A Prong 1: YES. Mental process capable of being performed by evaluation, opinion, and judgment.].
Claim 10
The computerized system of claim 1, wherein the method of the computerized system further comprises:
receiving an indication that a field assignee has input an electronic signature at the field of the document; and [Step 2A Prong 2: NO. Data inputting is insignificant extra-solution activity. The type or source of data does not cause the data inputting to integrate the exception into a practical application. Step 2B: NO. Under 2B this insignificant extra solution activity is well understood routine and conventional activity. See “Receiving or transmitting data over a network, e.g., using the Internet to gather data, Symantec, 838 F.3d at 1321, 120 USPQ2d at 1362.”]
in response to the receiving of the indication, automatically causing presentation at the document of the electronic signature such that the first user can view the electronic signature in near real-time relative to when the field assignee input the electronic signature at the field [Step 2A Prong 2: NO. Data outputting is insignificant extra-solution activity. The type or source of data does not cause the data outputting to integrate the exception into a practical application. Step 2B: NO. Under 2B this insignificant extra solution activity is well understood routine and conventional activity. See “Receiving or transmitting data over a network, e.g., using the Internet to gather data, Symantec, 838 F.3d at 1321, 120 USPQ2d at 1362.”].
Claim 11
The computerized system of claim 1, wherein the method of the computerized system further comprises:
in response to the generation of the field, receiving an indication that the first user has selected a field type for the field; and [Step 2A Prong 2: NO. Data inputting is insignificant extra-solution activity. The type or source of data does not cause the data inputting to integrate the exception into a practical application. Step 2B: NO. Under 2B this insignificant extra solution activity is well understood routine and conventional activity. See “Receiving or transmitting data over a network, e.g., using the Internet to gather data, Symantec, 838 F.3d at 1321, 120 USPQ2d at 1362.”]
in response to the receiving of the indication, automatically cause display, within the field, a string that describes the field type [Step 2A Prong 2: NO. Data outputting is insignificant extra-solution activity. The type or source of data does not cause the data outputting to integrate the exception into a practical application. Step 2B: NO. Under 2B this insignificant extra solution activity is well understood routine and conventional activity. See “Receiving or transmitting data over a network, e.g., using the Internet to gather data, Symantec, 838 F.3d at 1321, 120 USPQ2d at 1362.”].
Claim 12
A computer-implemented method comprising:
receiving one or more first natural language characters that were input by a first user at a document of an application, the document requiring at least one electronic signature by one or more entities, the one or more first natural language characters representing a first portion of a sentence; [Step 2A Prong 2: NO. Data inputting is insignificant extra-solution activity. The type or source of data does not cause the data inputting to integrate the exception into a practical application. Step 2B: NO. Under 2B this insignificant extra solution activity is well understood routine and conventional activity. See “Receiving or transmitting data over a network, e.g., using the Internet to gather data, Symantec, 838 F.3d at 1321, 120 USPQ2d at 1362.”]
in response to the receiving of the user input, providing the one or more first natural language characters as input into one or more machine learning models, wherein the one or more machine learning models [Step 2A Prong 2: NO. Data inputting is insignificant extra-solution activity. The type or source of data does not cause the data inputting to integrate the exception into a practical application. Recitation of generic computer components (e.g., machine learning models) is mere instruction to apply exception. Step 2B: NO. Under 2B this insignificant extra solution activity is well understood routine and conventional activity. See “Receiving or transmitting data over a network, e.g., using the Internet to gather data, Symantec, 838 F.3d at 1321, 120 USPQ2d at 1362.”] generates at least one of a field or one or more second natural language characters according to the one or more first natural language characters, the field being a data object representing a predetermined category for which a second user is to input data within the field according to the predetermined category, the one or more second natural language characters representing a second portion of the sentence, and the field representing a third portion of the sentence; and [Step 2A Prong 1: YES. Mental process capable of being performed by evaluation, opinion, and judgment.]
based at least in part on the generation, causing presentation of at least one of: the field next to the one or more first natural language characters in the document or the one or more second natural language characters next to the one or more first natural language characters at the document [Step 2A Prong 2: NO. Data outputting is insignificant extra-solution activity. The type or source of data does not cause the data outputting to integrate the exception into a practical application. Step 2B: NO. Under 2B this insignificant extra solution activity is well understood routine and conventional activity. See “Receiving or transmitting data over a network, e.g., using the Internet to gather data, Symantec, 838 F.3d at 1321, 120 USPQ2d at 1362.”].
Claim 13
The computer-implemented method of claim 12, wherein the one or more second natural language characters include at least one of:
a natural language sequence that represents a completed portion of the one or more first natural language characters of the user input, [Step 2A Prong 1: YES. Mental process capable of being performed by evaluation, opinion, and judgment.]
a template representing pre-formatted natural language content that was generated at a prior time or session relative to a time or session associated with the receiving of the one or more first natural language characters, or [Step 2A Prong 1: YES. Mental process capable of being performed by evaluation, opinion, and judgment.]
a name of an assignee of the field, wherein the assignee includes the second user [Step 2A Prong 1: YES. Mental process capable of being performed by evaluation, opinion, and judgment.].
Claim 14
The computer-implemented of claim 12, wherein the generation of at least one of: the one or more second natural language characters or the field is further based at least in part on a past history of computer inputs by the first user that input the one or more first natural language characters [Step 2A Prong 1: YES. Mental process capable of being performed by evaluation, opinion, and judgment. Step 2A Prong 2/Step 2B: NO. Recitation of generic computer components is mere instruction to apply the exception.].
Claim 15
The computer-implemented method of claim 12, further comprising:
causing presentation, at the document and at the user device, of a message that informs the first user of at least one of the one or more second natural language characters or the field is a suggestion to input next to a partial string representing the one or more first natural language characters; and [Step 2A Prong 2: NO. Data outputting is insignificant extra-solution activity. The type or source of data does not cause the data outputting to integrate the exception into a practical application. Step 2B: NO. Under 2B this insignificant extra solution activity is well understood routine and conventional activity. See “Receiving or transmitting data over a network, e.g., using the Internet to gather data, Symantec, 838 F.3d at 1321, 120 USPQ2d at 1362.”]
in response to receiving an indication of a user selection associated with the message, providing the user selection as feedback to the one or more machine learning models [Step 2A Prong 2: NO. Data inputting is insignificant extra-solution activity. The type or source of data does not cause the data inputting to integrate the exception into a practical application. Step 2B: NO. Under 2B this insignificant extra solution activity is well understood routine and conventional activity. See “Receiving or transmitting data over a network, e.g., using the Internet to gather data, Symantec, 838 F.3d at 1321, 120 USPQ2d at 1362.”].
Claim 16
The computer-implemented method of claim 12, further comprising;
at least partially in response to receiving the one or more first natural language characters, encoding the one or more first natural language characters into one or more word embedding feature vectors that represent positional information or context of each word in the one or more first natural language characters; and [Step 2A Prong 1: YES. Mental process capable of being performed by evaluation, opinion, and judgment.]
in response to the encoding, generating an attention vector for each word of the one or more first natural language characters by determining how relevant, via weighting, each word is relative to at least one other word and relative to at least one of the one or more second natural language characters and the field, and wherein the generating of at least one of the one or more second natural language characters or the field is based on the generation of an attention vector for each word [Step 2A Prong 1: YES. Mental process capable of being performed by evaluation, opinion, and judgment.].
Claim 17
The computer-implemented method of claim 12, further comprising: training or fine-tuning the one or more machine learning models by [Step 2A Prong 2/Step 2B: NO. Mere instruction to apply exception using generic computer components (e.g., machine learning models] learning string pairs or string-field pairs, each string-field pair indicates at field that is predicted to be placed next to a certain string, each string pair indicates a first string that is predicted to be placed next to a second string [Step 2A Prong 1: YES. Mental process capable of being performed by evaluation, opinion, and judgment.].
Claim 18
One or more non-transitory computer storage media having computer-executable instructions embodied thereon that, when executed, by one or more processors, cause the one or more processors to perform a method, the method comprising: [Step 1: YES. The claim is directed to a non-transitory computer storage media which is a statutory category.
receiving, over a computer network, a first request to open a document of an application, the document being native to the application, the document requiring at least one electronic signature by at least a first entity, the first request being issued at a user device associated with a user; [Step 2A Prong 2: NO. Data inputting is insignificant extra-solution activity. The type or source of data does not cause the data inputting to integrate the exception into a practical application. Step 2B: NO. Under 2B this insignificant extra solution activity is well understood routine and conventional activity. See “Receiving or transmitting data over a network, e.g., using the Internet to gather data, Symantec, 838 F.3d at 1321, 120 USPQ2d at 1362.”]
causing generation of a first partial string based on computer user input of a first partial string at a first page of the document, the first partial string being a first portion of a first sentence of an agreement that requires the at least one electronic signature by the first entity; [Step 2A Prong 1: YES. Mental process capable of being performed by evaluation, opinion, and judgment.]
causing generation of at least one of a first field or a second partial string based at least in part on one of, computer user input or a language model, the second partial string and the first field being a second portion of the first sentence of the agreement that requires the at least one electronic signature by at least the first entity; [Step 2A Prong 1: YES. Mental process capable of being performed by evaluation, opinion, and judgment.]
receiving, over the computer network and at the user device via a selection at the first page, a request to input a signature field at the first page of the agreement that requires the at least one electronic signature; [Step 2A Prong 2: NO. Data inputting is insignificant extra-solution activity. The type or source of data does not cause the data inputting to integrate the exception into a practical application. Step 2B: NO. Under 2B this insignificant extra solution activity is well understood routine and conventional activity. See “Receiving or transmitting data over a network, e.g., using the Internet to gather data, Symantec, 838 F.3d at 1321, 120 USPQ2d at 1362.”]
assigning the signature field to at least the first entity, the signature field being a data object for which at least the first entity is to input a signature within the signature field; and [Step 2A Prong 1: YES. Mental process capable of being performed by evaluation, opinion, and judgment.]
in response to the receiving of the request, automatically causing generation, at the first page of the document, of the signature field below the first partial string and at least one of the second partial string or the first field [Step 2A Prong 2: NO. Data outputting is insignificant extra-solution activity. The type or source of data does not cause the data outputting to integrate the exception into a practical application. Step 2B: NO. Under 2B this insignificant extra solution activity is well understood routine and conventional activity. See “Receiving or transmitting data over a network, e.g., using the Internet to gather data, Symantec, 838 F.3d at 1321, 120 USPQ2d at 1362.”].
Claim 19
The one or more non-transitory computer storage media of claim 18, wherein the language model generates the second partial string or the first field, and wherein the second partial string or field includes one of: [Step 2A Prong 1: YES. Mental process capable of being performed by evaluation, opinion, and judgment.]
a template representing pre-formatted natural language content that was generated at a prior time or session relative to a time or session associated with the receiving of the request; [Step 2A Prong 1: YES. Mental process capable of being performed by evaluation, opinion, and judgment.]
a natural language name of an assignee of the field, wherein the assignee includes the first entity, or [Step 2A Prong 1: YES. Mental process capable of being performed by evaluation, opinion, and judgment.]
a natural language name of the first field; [Step 2A Prong 1: YES. Mental process capable of being performed by evaluation, opinion, and judgment.].
Claim 20
The one or more non-transitory computer storage media of claim 18, the method of the one or more non-transitory computer storage media further comprising;
at least partially in response to receiving the one or more first natural language characters, encoding the one or more first natural language characters into one or more word embedding feature vectors that represent positional information or context of each word in the one or more first natural language characters; and [Step 2A Prong 1: YES. Mental process capable of being performed by evaluation, opinion, and judgment.]
in response to the encoding, generating an attention vector for each word of the one or more first natural language characters by determining how relevant, via weighting, each word is relative to at least one other word and relative to each word in the second partial string, and wherein the generating of the second partial string is based on the generation of an attention vector for each word [Step 2A Prong 1: YES. Mental process capable of being performed by evaluation, opinion, and judgment.].
Claim Rejections - 35 USC § 102
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 the appropriate paragraphs of 35 U.S.C. 102 that form the basis for the rejections under this section made in this Office action:
A person shall be entitled to a patent unless –
(a)(1) the claimed invention was patented, described in a printed publication, or in public use, on sale, or otherwise available to the public before the effective filing date of the claimed invention.
Claims 1, 3, 6, and 9 are rejected under 35 U.S.C. 102(a)(1) as being anticipated by Walters (US 2023/0138763 A1; published May 4, 2023).
Regarding claim 1, Walters discloses [a] computerized system comprising:
one or more processors; and (Walters ¶ 43 (“Consistent with disclosed embodiments, computing resources 103 may comprise one or more processors and one or more memories. A processor (or processors) can be one or more data or software processing devices. For example, the processor may take the form of, but is not limited to, a microprocessor, embedded processor, or the like, or may be integrated in a system on a chip (SoC). Furthermore, according to some embodiments, the processor may be a processor manufactured by Intel®, AMD®, Qualcomm®, Apple®, NVIDIA®, or the like.”))
computer storage memory having computer-executable instructions stored thereon which, when executed by the one or more processors, implement a method comprising: (Walters ¶ 44 (“A memory (or memories) may include one or more storage devices configured to store instructions used by the processor to perform functions related to disclosed embodiments. Memory may be configured to store software instructions, such as programs, that when executed by processor, perform one or more operations consistent with disclosed embodiments.”))
receiving, over a computer network, a first request to generate one or more elements of a document of a web application, the document being native to the web application, the document requiring at least one electronic signature by one or more entities, the first request being issued at a user device associated with a first user; and (Walters ¶ 71 (“At step 401, process 400 may include receiving a request for synthetic data [e.g., elements]. A request for synthetic data may be received from for example, a computing device [e.g., user device] associated with a user [e.g., first user] or through an interface provided to a user.”), ¶ 72 (“As described herein, synthetic data may take a variety of formats. For example, a synthetic data format may be a text-based format. Text-based formats include, but are not limited to, documents, forms, written communications, articles, and the like. In some embodiments, a synthetic data format may indicate a document type, such as a contract [e.g., the document requiring at least one electronic signature by one or more entities], email, patent, application, etc. Disclosed embodiments are not limited to text-based formats of synthetic data. For example, synthetic data formats may be audio, image, or video formats.”))
based at least in part on at least one of: the first request or user input of one or more natural language characters at the document, automatically causing generation, via a machine learning model, of at least one of: (Walters ¶ 73 (“At step 403, process 400 may include selecting, based on the synthetic data format, a first syntax generator [e.g., machine learning model]. A syntax generator may be selected from, for example, syntax generator module 107. As described above, syntax generator module 107 may store a variety of syntax generator models. The syntax generator models may each be trained to generate a specific syntax for a certain data format (e.g., emails, documents, images, etc.). Accordingly, a specific syntax generator model may be selected based on the desired format of the synthetic data ultimately to be produced.”))
(1) one or more strings at the document, or (Walters ¶ 81 (“At step 405, process 400 may include generating, using the first syntax generator, a token set. As described above, a token set may comprise one or more tokens. Tokens may correspond to various types of words or information. Accordingly tokens may have a corresponding token type that indicates the type of word or information that the token represents. For example, a token may have a type corresponding to a filler word [e.g., strings], a name, address, account number, social security number, part of speech, color, sound, image, or word or information type. The token type may be based on a data type and/or data format identified at step 403.”), ¶ 93 (“Step 411 of process 400 may include generating data content. The selected content generators may be executed to generate data content corresponding to tokens of the token set generated by the syntax generator at step 405. As described herein, each selected content generator may generate data content corresponding to a different token type of the token set.”))
(2) a field, the field being a data object representing a predetermined category for which a second user is to input data within the field according to the predetermined category (Walters ¶ 81 (“At step 405, process 400 may include generating, using the first syntax generator, a token set. As described above, a token set may comprise one or more tokens. Tokens may correspond to various types of words or information. Accordingly tokens may have a corresponding token type that indicates the type of word or information that the token represents. For example, a token may have a type corresponding to a filler word, a name, address, account number, social security number [e.g., field], part of speech, color, sound, image, or word or information type. The token type [e.g., predetermined category] may be based on a data type and/or data format identified at step 403.”), ¶ 93 (“Step 411 of process 400 may include generating data content. The selected content generators may be executed to generate data content corresponding to tokens of the token set generated by the syntax generator at step 405. As described herein, each selected content generator may generate data content corresponding to a different token type of the token set.”)).
Regarding claim 3, Walters discloses the invention of claim 1 as discussed above. Walters further discloses wherein the generation of at least one of: the one or more strings or the field is further based at least in part on a past history of computer inputs by the first user that input the one or more natural language characters (Walters ¶ 100 (“For example, process 500 may be used to train and refine generators to produce synthetic data that falls within a certain range of relative similarity to a reference data set, according to a calculated similarity metric. In some cases, there may be a range of acceptable similarity metrics to the reference data set for new synthetic data. For example, generated synthetic data that is identical or nearly identical to a reference data set may not be useful for training other models because it may not have sufficient variation from the reference data set to meaningfully train the model. However, synthetic data that falls below a threshold of similarity may also not be useful in some cases. For example, if the reference data set is an email, but the generated synthetic data does not resemble an email, the synthetic data may not be useful for, as an example, training another model using synthetic emails.”)).
Regarding claim 6, Walters discloses the invention of claim 1 as discussed above. Walters further discloses: training or fine-tuning the machine learning model by learning string pairs or string-field pairs, each string-field pair indicates at field that is predicted to be placed next to a certain string, each string pair indicates a first string that is predicted to be placed next to a second string (Walters ¶ 14 (“The method may include generating a token set comprising a first token and a second token using the first syntax generator and identifying a first token type corresponding to the first token and a second token type corresponding to the second token.”), ¶ 104 (“At step 503, process 500 may include determining whether the similarity metric meets a similarity threshold. If the similarity metric meets the threshold (e.g., is not too similar or not too dissimilar), process 500 may proceed step 505 and return the generated synthetic data without refining any of the models. If the similarity metric fails to meet the threshold (e.g., the synthetic data is either too similar or too dissimilar to the reference data), process 500 may proceed to step 507. At step 507, process 500 may include retraining one or more of a syntax generator or content generator used to generate the synthetic data. For example, if data are not similar enough, process 500 may include retraining or refining the syntax generator used to generate the syntax-token set for the synthetic data. Retraining or refining may include adjusting parameter values, adjusting a number of layers, changing a model type, adjusting model weights, or making other suitable changes to a model. At step 509, process 500 may include using the retrained generator or generators to generate a second set of synthetic data. In some embodiments, process 500 may then proceed back to step 501, where it may include calculating a similarity metric for the second set of synthetic data, as described above.”)).
Regarding claim 9, Walters discloses the invention of claim 1 as discussed above. Walters further discloses wherein the field is one of: a signature field, a company name field, a personal name field, an email address field, a job title field, or a residential or business address field (Walters ¶ 46 (“The syntax may indicate for example, the types of words or information included in the sentence, phrase, document, communication, etc. For example, the syntax may include indications of different parts of speech, names, addresses, numbers, colors, images, sounds, or other suitable information.”)).
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.
The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or non-obviousness.
Claim 2 is rejected under 35 U.S.C. 103 as being unpatentable over Walters as applied to claim 1 above, and further in view of Bower (US 2008/0195388 A1; published Aug. 14, 2008).
Regarding claim 2, Walters discloses the invention of claim 1 as discussed above. Walters further discloses wherein the one or more strings include at least one of:
a natural language sequence (Walters ¶ 72 (“As described herein, synthetic data may take a variety of formats. For example, a synthetic data format may be a text-based format. Text-based formats include, but are not limited to, documents, forms, written communications, articles, and the like. In some embodiments, a synthetic data format may indicate a document type, such as a contract, email, patent, application, etc. Disclosed embodiments are not limited to text-based formats of synthetic data. For example, synthetic data formats may be audio, image, or video formats.”)).
Walters does not expressly disclose that the synthetic text data represents a completed portion of the one or more natural language characters of the user input, (but see Bower ¶ 51 (“As the user begins entering reply text or data, the input method 115 intercepts each character of entered data on a character-by-character basis at operation 345. At operation 350, the input method 115 calls the prediction engine 125 to obtain prediction results responsive to the entered text or data character. At operation 360, the prediction engine 125 [e.g., language model] obtains words from both the application defined data source 150 via the application defined candidate provider 155 and from the existing text prediction data sources 135 via the static word provider 140 at operation 360.”)).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified Walters to incorporate the teachings of Bower to complete the input text, at least because typing or otherwise entering information into a computing device can be cumbersome and time consuming where each individual word must be typed in its entirety. See Bower ¶ 1.
Walters further discloses:
a template representing pre-formatted natural language content that was generated at a prior time or session relative to a time or session associated with the receiving of the first request, or (Walters ¶ 73 (“At step 403, process 400 may include selecting, based on the synthetic data format, a first syntax generator. A syntax generator may be selected from, for example, syntax generator module 107. As described above, syntax generator module 107 may store a variety of syntax generator models. The syntax generator models may each be trained to generate a specific syntax for a certain data format (e.g., emails, documents, images, etc.). Accordingly, a specific syntax generator model may be selected based on the desired format of the synthetic data ultimately to be produced.”), ¶ 77 (“As an example, various syntax generators may be configured to generate syntax for a sentence. Accordingly, the tokens within the token sets produced by such syntax generators may relate to, for example, certain parts of speech (e.g., noun, verbs, adjectives, adverbs, conjunctions, etc.). In such a case, the sentence syntax may be language specific. For example, English and Spanish may have different placement of articles and adjectives, requiring different orders of the corresponding tokens in a token set to generate realistic sentences.”))
a name of an assignee of the field, wherein the assignee includes the second user (Walters ¶ 81 (“At step 405, process 400 may include generating, using the first syntax generator, a token set. As described above, a token set may comprise one or more tokens. Tokens may correspond to various types of words or information. Accordingly tokens may have a corresponding token type that indicates the type of word or information that the token represents. For example, a token may have a type corresponding to a filler word, a name, address, account number, social security number, part of speech, color, sound, image, or word or information type. The token type may be based on a data type and/or data format identified at step 403.”)).
Claim 4 is rejected under 35 U.S.C. 103 as being unpatentable over Walters as applied to claim 1 above, and further in view of Gupta (US 2014/0181692 A1; published Jun. 26, 2014).
Regarding claim 4, Walters discloses the invention of claim 1 as discussed above. Walters does not expressly disclose causing presentation, at the document and at the user device, of a message that informs the first user that the at least one of the one or more strings or the field is a suggestion to input next to a partial string representing the one or more natural language characters (but see Gupta ¶ 4 (“Embodiments are provided for persisting atomically linked entities when utilizing an auto-complete mechanism. A computing device may be utilized to receive an input in a user interface. The computing device may then display an auto-complete suggestion list in response to receiving the input in the user interface. A selection of an entity may then be received from the auto-complete suggestion list. The selected entity may then be atomically linked to program code defining an action. The atomically linked entity may then be inserted in the user interface. The atomically linked entity may then be persisted among the input received within the user interface. The input is modifiable and the atomically linked entity is unmodifiable within the user interface.”)).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified Walters to incorporate the teachings of Gupta to display suggested auto-completions after an input received at a user interface, at least because doing so would facilitate the posting of content.
Claim 5 is rejected under 35 U.S.C. 103 as being unpatentable over Walters as applied to claim 1 above, and further in view of Shazeer (US 2018/0341860 A1; published Nov. 29, 2018).
Regarding claim 5, Walters discloses the invention of claim 1 as discussed above. Walters does not expressly disclose: wherein the method of the computerized system further comprising;
at least partially in response to receiving the user input of the one or more natural language characters at the document, encoding the one or more natural language characters into one or more word embedding feature vectors that represent positional information or context of each word in the user input; and (but see Shazeer ¶ 29 (“The embedding layer 120 is configured to, for each network input in the input sequence, map the network input to a numeric representation of the network input in an embedding space, e.g., into a vector in the embedding space. The embedding layer 120 then provides the numeric representations of the network inputs to the first subnetwork in the sequence of encoder subnetworks 130, i.e., to the first encoder subnetwork 130 of the N encoder subnetworks 130.”))
in response to the encoding, generating an attention vector for each word of the user input by determining how relevant, via weighting, each word in the user input is relative to at least one other word in the user input and the one of the one or more strings and the field, and wherein the generating of at least one of the one or more strings or the field is based on the generation of an attention vector for each word (but see Shazeer ¶ 30 (“In particular, in some implementations, the embedding layer 120 is configured to map each network input to an embedded representation of the network input and then combine, e.g., sum or average, the embedded representation of the network input with a positional embedding of the input position of the network input in the input order to generate a combined embedded representation of the network input. That is, each position in the input sequence has a corresponding embedding and for each network input the embedding layer 120 combines the embedded representation of the network input with the embedding of the network input's position in the input sequence. Such positional embeddings can enable the model to make full use of the order of the input sequence without relying on recurrence or convolutions.”)).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified Walters to incorporate the teachings of Shazeer to encode and embed the words and other information to generate synthetic data, at least because doing so would enable creating content for multiple different contexts associated with different syntax generators. See Walters ¶ 35.
Claim 7 is rejected under 35 U.S.C. 103 as being unpatentable over Walters as applied to claim 1 above, and further in view of Bedi (US 2022/0114326 A1; published Apr. 14, 2022).
Regarding claim 7, Walters discloses the invention of claim 1 as discussed above. Walters does not expressly disclose:
as part of the automatic generation of at least one of the one or more strings or the field, automatically moving at least one of: the one or more natural language characters, the one or more strings, or the field from a first line to a second line in the document based at least in part on a screen size of the user device (but see Bedi ¶ 20 (“Given an input layout and a target layout size, this disclosed system aims to create different layout variations with the target layout size based on the design elements extracted from the input layout. Specifically, the disclosed technical solution enables the automatic generation of the output design variations based on the visual flow of design blocks in the input layout. Accordingly, the disclosed technologies allow designers to reproduce layout variations as needed for different mediums or devices in a fast and efficient manner.”), ¶ 43 (“The poster includes both textual and graphic elements to be both eye-catching and informative, although other posters may be either wholly graphical or wholly text [e.g., natural language characters].”), ¶ 58 (“For an allotted pattern, the system uses the visual flow information to adjust a pattern (e.g., the demarcations), then to reposition, resize, rescale, or reflow individual design blocks or design elements to transform them into the pattern. Notably, the respective ranks of design blocks are used by the system to decide how to reposition, resize, rescale, or reflow a design block in the graphic design.”) (reflowing text refers to the repositioning of text based on the size of the display screen and includes moving the text from one line to another line in order to fit the text on the display)).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified Walters to incorporate the teachings of Bedi to reflow the synthetic data generated by the syntax generators based on the display size, at least because doing so “enables the automatic generation of the output design variations based on the visual flow of design blocks in the input layout.” Bedi ¶ 20.
Claim 8 is rejected under 35 U.S.C. 103 as being unpatentable over Walters as applied to claim 1 above, and further in view of Gonser (US 2008/0209313 A1; published Aug. 28, 2008).
Regarding claim 8, Walters discloses the invention of claim 1 as discussed above. Walters does not expressly disclose:
subsequent to the automatic generation of at least one of the one or more strings or the field, assigning the field to the second user based on second computer user input from the first user, the assigning being indicative of authorizing only the second user, and not the first user, to populate the field; and (but see Gonser ¶ 19 (“At block 44 signature locations and data fields are assigned in the data template. Signature fields are located where a party is asked to sign the document. Once the area is located a tab is entered to direct a party to the transaction to that location. Locations where a party is to initial are assigned as well as data fields such as date signed and printed name of the signer. In one embodiment the data fields are automatically entered by the system to ensure that the correct date is entered. At block 46 the template data is bound to the document. The template information as entered by the party and the document to be signed are merged. At block 48 the digital document, including the template, are sent to all of the parties to the transaction for signing.”))
in response to the assigning, automatically causing a transmission, over the computer network, of an indication to a device associated with the second user, wherein the second user is able to populate the field, at the document, based on the assigning of the field and the transmission of the indication (but see Gonser ¶ 19 (“At block 46 the template data is bound to the document. The template information as entered by the party and the document to be signed are merged. At block 48 the digital document, including the template, are sent to all of the parties to the transaction for signing.”)).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified Walters to incorporate the teachings of Gonser to assign the name, address, signature fields to users and transmit the document to the parties to the transaction for signing, at least because doing so would enable signing, storing, and routing legal documents to the necessary location in a controllable fashion. See Gonser ¶ 2.
Claim 10 is rejected under 35 U.S.C. 103 as being unpatentable over Walters as applied to claim 1 above, and further in view of Hancock (US 2019/0074976 A1; published Mar. 7, 2019).
Regarding claim 10, Walters discloses the invention of claim 1 as discussed above. Walters does not expressly disclose:
receiving an indication that a field assignee has input an electronic signature at the field of the document; and (but see Hancock ¶ 27 (“In an example, the signatures of the signees 101 a-b is captured by touch screen, pre-scanned image, mouse or signature pad.”))
in response to the receiving of the indication, automatically causing presentation at the document of the electronic signature such that the first user can view the electronic signature in near real-time relative to when the field assignee input the electronic signature at the field (but see Hancock ¶ 25 (“In an example, the collaborative signature system includes at least one template (see, e.g., FIG. 2) stored in a database 140, the at least one template representing a document that is to be signed. The example collaborative signature system also includes a remote server 110 executing program code 150 to deliver the at least one template to at least a first signee 101 a and a second signee 101 b so that both signees can see the document to be signed substantially simultaneously. The remote server 110 processes and displays for both signees 101 a and 101 b the signature of the first signee substantially simultaneously. The remote server 110 also processes and displays for both signees 101 a-b the signature of the second signee substantially simultaneously.”), ¶ 26 (“The remote server may also process and display for both signees substantially simultaneously other entries to the document to be signed. For example, the other entries include at least one checkbox selectable by at least one of the signees. The other entries may also include at least one text input by at least one of the signees.”)).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified Walters to incorporate the teachings of Hancock to simultaneously display signatures of either party, at least because doing so would enable the document to be signed immediately without being in the presence of both parties at the same time.
Claim 11 is rejected under 35 U.S.C. 103 as being unpatentable over Walters as applied to claim 1 above, and further in view of Kopikare (US 2018/0089412 A1; published Mar. 29, 2018).
Regarding claim 11, Walters discloses the invention of claim 1 as discussed above. Walters does not expressly disclose: in response to the generation of the field, receiving an indication that the first user has selected a field type for the field; and in response to the receiving of the indication, automatically cause display, within the field, a string that describes the field type. However, Kopikare teaches providing a set of formatting controls from which a survey administrator may choose a signature text control for inclusion in an electronic survey. Paragraph 63. Kopikare further teaches adding a signature text control within an electronic survey, within the signature text control the text box shows text entered in by the survey administrator. Paragraph 65. As a default, the text box states, “Sign Here,” and is displayed within the signature field above the signature line. Id.
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified Matveyenko to incorporate the teachings of Kopikare to automatically display a default description of a control added to a form, at least because doing so would assist a user filling out the form.
Claims 12-14 and 16 are rejected under 35 U.S.C. 103 as being unpatentable over Shazeer, in view of Walters and Zhang (US 2021/0319178 A1; published Oct. 14, 2021).
Regarding claim 12, Shazeer discloses [a] computer-implemented method comprising:
receiving one or more first natural language characters that were input by a first user at a document of an application…the one or more first natural language characters representing a first portion of a sentence; (Shazeer ¶ 83 (“The system receives an input sequence (step 310).”), ¶ 17 (“As another example, the system may be a natural language processing system. For example, if the input sequence is a sequence of words in an original language, e.g., a sentence or phrase,…”)).
Shazeer does not expressly disclose the document requiring at least one electronic signature by one or more entities (but see Walters ¶ 46 (“Environment 101 may also include a syntax generator module 107. Syntax generator module 107 may be a storage device configured to store one or more syntax generators. Syntax generator module 107 may also store information related to various syntax generators, for example, indicating a type of syntax or data corresponding to a particular syntax generator. As used herein, a syntax generator is a data model configured to generate syntax for synthetic data. Syntax for synthetic data includes, but is not limited to, format or structure of synthetic data according to one or more rules. For example, syntax of synthetic data may be in the form of a sentence or phrase. As another example, syntax of synthetic data may be a particular type of document or communications, such as a contract, patent, email, letter, text message, or others. [e.g., document requiring at least one electronic signature by one or more entities]”)).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified Shazeer to incorporate the teachings of Walters to generate synthetic content for a particular type of document, e.g., a contract, at least because doing so would enable systems and methods of creating synthetic data that can be used across varying problem domains. See Walters ¶ 3.
Shazeer further discloses:
in response to the receiving of the user input, providing the one or more first natural language characters as input into one or more machine learning models, wherein the one or more machine learning models generates at least one of a field or one or more second natural language characters according to the one or more first natural language characters, the field being a data object representing a predetermined category for which a second user is to input data within the field according to the predetermined category, the one or more second natural language characters representing a second portion of the sentence, and the field representing a third portion of the sentence; and (Shazeer ¶ 17 (“the output sequence may be a summary of the input sequence in the original language, i.e., a sequence that has fewer words than the input sequence but that retains the essential meaning of the input sequence. As another example, if the input sequence is a sequence of words that form a question, the output sequence can be a sequence of words that form an answer to the question.”), ¶ 4 (“In particular, the system generates the output sequence using an encoder neural network and a decoder neural network that are both attention-based.”)).
Shazeer does not expressly disclose:
based at least in part on the generation, causing presentation of at least one of: the field next to the one or more first natural language characters in the document or the one or more second natural language characters next to the one or more first natural language characters at the document (but see Zhang Abstract (“There is provided a computer implemented method of context based autocomplete of text, comprising: receiving input text, feeding the input text into a context-prediction component of a machine learning model that predicts a certain context of a plurality of contexts, selecting a certain context-specific component of the machine learning model from a plurality of context-specific components according to the certain context, feeding the input text into the selected context-specific component that outputs autocomplete text, and providing the autocomplete text.”)).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified Shazeer to incorporate the teachings of Zhang to use the transformer model to generate and output autocomplete text, at least because doing so would enable autocomplete is a software feature that predicts the rest of a word, sentence, or paragraph a user is typing. Zhang ¶ 2.
Regarding claim 13, Shazeer, in view of Walters and Zhang, discloses the invention of claim 12 as discussed above. Shazeer further discloses wherein the one or more second natural language characters include at least one of:
a natural language sequence that represents a completed portion of the one or more first natural language characters of the user input, (Shazeer ¶ 17 (“the output sequence may be a summary of the input sequence in the original language, i.e., a sequence that has fewer words than the input sequence but that retains the essential meaning of the input sequence. As another example, if the input sequence is a sequence of words that form a question, the output sequence can be a sequence of words that form an answer to the question.”), ¶ 4 (“In particular, the system generates the output sequence using an encoder neural network and a decoder neural network that are both attention-based.”)).
a template representing pre-formatted natural language content that was generated at a prior time or session relative to a time or session associated with the receiving of the one or more first natural language characters, or
a name of an assignee of the field, wherein the assignee includes the second user.
Regarding claim 14, Shazeer, in view of Walters and Zhang, discloses the invention of claim 12 as discussed above. Shazeer further discloses wherein the generation of at least one of: the one or more second natural language characters or the field is further based at least in part on a past history of computer inputs by the first user that input the one or more first natural language characters (Shazeer ¶ 14 (“This specification describes a system implemented as computer programs on one or more computers in one or more locations that generates an output sequence that includes a respective output at each of multiple positions in an output order from an input sequence that includes a respective input at each of multiple positions in an input order, i.e., transduces the input sequence into the output sequence.”)).
Regarding claim 16, Shazeer, in view of Walters and Zhang, discloses the invention of claim 12 as discussed above. Shazeer further discloses:
at least partially in response to receiving the one or more first natural language characters, encoding the one or more first natural language characters into one or more word embedding feature vectors that represent positional information or context of each word in the one or more first natural language characters; and (Shazeer ¶ 29 (“The embedding layer 120 is configured to, for each network input in the input sequence, map the network input to a numeric representation of the network input in an embedding space, e.g., into a vector in the embedding space. The embedding layer 120 then provides the numeric representations of the network inputs to the first subnetwork in the sequence of encoder subnetworks 130, i.e., to the first encoder subnetwork 130 of the N encoder subnetworks 130.”))
in response to the encoding, generating an attention vector for each word of the one or more first natural language characters by determining how relevant, via weighting, each word is relative to at least one other word and relative to at least one of the one or more second natural language characters and the field, and wherein the generating of at least one of the one or more second natural language characters or the field is based on the generation of an attention vector for each word (Shazeer ¶ 30 (“In particular, in some implementations, the embedding layer 120 is configured to map each network input to an embedded representation of the network input and then combine, e.g., sum or average, the embedded representation of the network input with a positional embedding of the input position of the network input in the input order to generate a combined embedded representation of the network input. That is, each position in the input sequence has a corresponding embedding and for each network input the embedding layer 120 combines the embedded representation of the network input with the embedding of the network input's position in the input sequence. Such positional embeddings can enable the model to make full use of the order of the input sequence without relying on recurrence or convolutions.”)).
Claim 15 is rejected under 35 U.S.C. 103 as being unpatentable over Shazeer, Walters, and Zhang as applied to claim 12 above, and further in view of Gupta and Sahuguet (US 2016/0041991 A1; published Feb. 11, 2016).
Regarding claim 15, Shazeer, in view of Walters and Zhang, discloses the invention of claim 12 as discussed above. Shazeer does not expressly disclose:
causing presentation, at the document and at the user device, of a message that informs the first user of at least one of the one or more second natural language characters or the field is a suggestion to input next to a partial string representing the one or more first natural language characters; and (but see Gupta (US 2014/0181692 A1; published Jun. 26, 2014) ¶ 4 (“Embodiments are provided for persisting atomically linked entities when utilizing an auto-complete mechanism. A computing device may be utilized to receive an input in a user interface. The computing device may then display an auto-complete suggestion list in response to receiving the input in the user interface. A selection of an entity may then be received from the auto-complete suggestion list. The selected entity may then be atomically linked to program code defining an action. The atomically linked entity may then be inserted in the user interface. The atomically linked entity may then be persisted among the input received within the user interface. The input is modifiable and the atomically linked entity is unmodifiable within the user interface.”)).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified Walters to incorporate the teachings of Gupta to display suggested auto-completions after an input received at a user interface, at least because doing so would facilitate the posting of content.
Shazeer does not expressly disclose in response to receiving an indication of a user selection associated with the message, providing the user selection as feedback to the one or more machine learning models (but see Sahuguest ¶ 42 (“Next, a selection of a query autocomplete, e.g., “restaurant with terrace,” may be received (block 620). Query results may be provided (block 622) based on the selected search query of “restaurant with terrace.” Additionally, as described above, the selection and associated environmental context may be stored in the query logs 626 to provide feedback to the query autocomplete determinations. Thus, the selected query autocomplete of “restaurant with terrace” and the current weather may be stored (block 624), such as in the query logs (block 626). The selection may thus increase popularity of the query autocomplete “restaurant with terrace” for the environmental context of current weather.”)).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified Shazeer to incorporate the teachings of Sahuguet to provide feedback to the autoencoder, at least because doing so would improve the model.
Claim 17 is rejected under 35 U.S.C. 103 as being unpatentable over Shazeer, Walters, and Zhang as applied to claim 12 above, and further in view of Florencio (US 2021/0133438 A1; published May 6, 2021).
Regarding claim 17, Shazeer, in view of Walters and Zhang, discloses the invention of claim 12 as discussed above. Shazeer does not expressly disclose training or fine-tuning the one or more machine learning models by learning string pairs or string-field pairs, each string-field pair indicates at field that is predicted to be placed next to a certain string, each string pair indicates a first string that is predicted to be placed next to a second string (but see Florencio Abstract (“Interfaces and systems are provided for harvesting ground truth from forms to be used in training models based on key-value pairings in the forms and to later use the trained models to identify related key-value pairings in new forms. Initially, forms are identified and clustered to identify a subset of forms to label with the key-value pairings. Users provide input to identify keys to use in labeling and then select/highlight text from forms that are presented concurrently with the keys in order to associate the highlighted text with the key(s) as the corresponding key-value pairing(s). After labeling the forms with the key-value pairings, the key-value pairing data is used as ground truth for training a model to independently identify the key-value pairing(s) in new forms. Once trained, the model is used to identify the key-value pairing(s) in new forms.”)).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified Walters to incorporate the teachings of Florencio to train the syntax generators to identify key-value pairs in new forms, at least because doing so would enable applications to parse forms to identify desired content within the forms. See Florencio ¶ 4.
Claim 18 is rejected under 35 U.S.C. 103 as being unpatentable over Bower, in view of Dunn (US 2011/0093807 A1; published Apr. 21, 2011) and Gonser.
Regarding claim 18, Bower discloses [o]ne or more non-transitory computer storage media having computer-executable instructions embodied thereon that, when executed, by one or more processors, cause the one or more processors to perform a method, the method comprising: (Bower ¶ 78 (“By way of example, and not limitation, computer-readable media may comprise computer storage media and communication media. Computer storage media includes volatile and non-volatile, removable and non-removable media implemented in any method or technology for storage of information such as computer-readable instructions, data structures, program modules or other data.”))
receiving, over a computer network, a first request to open a document of an application, the document being native to the application, the document requiring at least one electronic signature by at least a first entity, the first request being issued at a user device associated with a user; (Bower ¶ 49 (“FIG. 3 is a state diagram and operational flow illustrating a method for providing context-based word prediction via an electronic messaging application. The operational flow and components illustrated in FIG. 3 provide further detail with respect to operation of embodiments of the present invention with respect to an electronic mail messaging application. Referring to FIG. 3, the electronic mail messaging application 170 [e.g., application] begins at operation 310 when a user 110 starts a reply action [first request] by attempting to reply to a previously received electronic mail message (see Example 1 above). At operation 315, the messaging application 170 creates a new reply window in which the user may type or otherwise enter a reply electronic mail message [e.g., document] (see Example 2 above).”))
causing generation of a first partial string based on computer user input of a first partial string at a first page of the document, the first partial string being a first portion of a first sentence of an agreement that requires the at least one electronic signature by the first entity; (Bower ¶ 51 (“Referring back to the messaging application 170, at operation 330, the user 110 begins typing or otherwise entering text or data [e.g., computer user input of a first partial string] into the reply message window in response to the received electronic mail message.”),
causing generation of at least one of a first field or a second partial string based at least in part on one of, computer user input or a language model, the second partial string and the first field being a second portion of the first sentence of the agreement that requires the at least one electronic signature by at least the first entity; (Bower ¶ 51 (“As the user begins entering reply text or data, the input method 115 intercepts each character of entered data on a character-by-character basis at operation 345. At operation 350, the input method 115 calls the prediction engine 125 to obtain prediction results responsive to the entered text or data character. At operation 360, the prediction engine 125 [e.g., language model] obtains words from both the application defined data source 150 via the application defined candidate provider 155 and from the existing text prediction data sources 135 via the static word provider 140 at operation 360.”)).
Bower does not expressly disclose:
receiving, over the computer network and at the user device via a selection at the first page, a request to input a signature field at the first page of the agreement that requires the at least one electronic signature; (but see Dunn ¶ 65 (“FIG. 11D shows a screen in which the document sender selected two signer users ("Jonathan Siegel" and "Jeff Siegel"), and also selected a non-signer user to receive a copy of the document, including the signed document ("cc" "Jones Siegel"). FIG. 11E shows a document sender selecting a signature entry pad as a document overlay to insert in a signature field in the electronic document. FIG. 11F shows the screen that allows the document sender to identify which signer user is associated with which signature field. The screen allows the document sender to indicate whether a signature is required or optional, and associates a name with a given signature field for easily inserting the signature field in multiple locations in the document. FIG. 11G shows the signature entry pad associated with the signature field identified in the previous screen, located at the desired location within the document.”)).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified Bower to incorporate the teachings of Dunn to insert a signature field in an agreement document, at least because doing so would facilitate completion of an electronic document via a signer user over the internet. Dunn ¶ 7.
Bower does not expressly disclose:
assigning the signature field to at least the first entity, the signature field being a data object for which at least the first entity is to input a signature within the signature field; and (but see Gonser ¶ 19 (“At block 44 signature locations and data fields are assigned in the data template. Signature fields are located where a party is asked to sign the document. Once the area is located a tab is entered to direct a party to the transaction to that location. Locations where a party is to initial are assigned as well as data fields such as date signed and printed name of the signer. In one embodiment the data fields are automatically entered by the system to ensure that the correct date is entered. At block 46 the template data is bound to the document. The template information as entered by the party and the document to be signed are merged. At block 48 the digital document, including the template, are sent to all of the parties to the transaction for signing.”)).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified Walters to incorporate the teachings of Gonser to assign the name, address, signature fields to users and transmit the document to the parties to the transaction for signing, at least because doing so would enable signing, storing, and routing legal documents to the necessary location in a controllable fashion. See Gonser ¶ 2.
Bower does not expressly disclose:
in response to the receiving of the request, automatically causing generation, at the first page of the document, of the signature field below the first partial string and at least one of the second partial string or the first field (but see Dunn ¶ 65 (“FIG. 11D shows a screen in which the document sender selected two signer users ("Jonathan Siegel" and "Jeff Siegel"), and also selected a non-signer user to receive a copy of the document, including the signed document ("cc" "Jones Siegel"). FIG. 11E shows a document sender selecting a signature entry pad as a document overlay to insert in a signature field in the electronic document. FIG. 11F shows the screen that allows the document sender to identify which signer user is associated with which signature field. The screen allows the document sender to indicate whether a signature is required or optional, and associates a name with a given signature field for easily inserting the signature field in multiple locations in the document. FIG. 11G shows the signature entry pad associated with the signature field identified in the previous screen, located at the desired location within the document.”); see FIG. 11G (illustrating signature field below the text entered by user)).
Bowers is combined with Dunn according to the same rationale as stated above.
Claim 19 is rejected under 35 U.S.C. 103 as being unpatentable over Bower, Dunn, and Gonser as applied to claim 18 above, and further in view of Walters.
Regarding claim 19, Bower, in view of Dunn and Gonser, discloses the invention of claim 18 as discussed above. Bowers further discloses wherein the language model generates the second partial string or the first field, (Bower ¶ 51 (“As the user begins entering reply text or data, the input method 115 intercepts each character of entered data on a character-by-character basis at operation 345. At operation 350, the input method 115 calls the prediction engine 125 to obtain prediction results responsive to the entered text or data character. At operation 360, the prediction engine 125 [e.g., language model] obtains words from both the application defined data source 150 via the application defined candidate provider 155 and from the existing text prediction data sources 135 via the static word provider 140 at operation 360.”)).
Bower does not expressly disclose and wherein the second partial string or field includes one of:
a template representing pre-formatted natural language content that was generated at a prior time or session relative to a time or session associated with the receiving of the request; (but see Walters ¶ 73 (“At step 403, process 400 may include selecting, based on the synthetic data format, a first syntax generator. A syntax generator may be selected from, for example, syntax generator module 107. As described above, syntax generator module 107 may store a variety of syntax generator models. The syntax generator models may each be trained to generate a specific syntax for a certain data format (e.g., emails, documents, images, etc.). Accordingly, a specific syntax generator model may be selected based on the desired format of the synthetic data ultimately to be produced.”), ¶ 77 (“As an example, various syntax generators may be configured to generate syntax for a sentence. Accordingly, the tokens within the token sets produced by such syntax generators may relate to, for example, certain parts of speech (e.g., noun, verbs, adjectives, adverbs, conjunctions, etc.). In such a case, the sentence syntax may be language specific. For example, English and Spanish may have different placement of articles and adjectives, requiring different orders of the corresponding tokens in a token set to generate realistic sentences.”))
a natural language name of an assignee of the field, wherein the assignee includes the first entity, or (but see Walters ¶ 81 (“At step 405, process 400 may include generating, using the first syntax generator, a token set. As described above, a token set may comprise one or more tokens. Tokens may correspond to various types of words or information. Accordingly tokens may have a corresponding token type that indicates the type of word or information that the token represents. For example, a token may have a type corresponding to a filler word, a name, address, account number, social security number, part of speech, color, sound, image, or word or information type. The token type may be based on a data type and/or data format identified at step 403.”))
a natural language name of the first field;
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified Bower to incorporate the teachings of Walters to generate a specific syntax for a certain data format, at least because doing so would enable creating synthetic data similar to existing datasets that can be used across varying problem domains. See Walters ¶ 3.
Claim 20 is rejected under 35 U.S.C. 103 as being unpatentable over Bower, Dunn, and Gonser as applied to claim 18 above, and further in view of Shazeer.
Regarding claim 20, Bower, in view of Dunn and Gonser, discloses the invention of claim 18 as discussed above. Bower does not expressly disclose:
at least partially in response to receiving the one or more first natural language characters, encoding the one or more first natural language characters into one or more word embedding feature vectors that represent positional information or context of each word in the one or more first natural language characters; and (but see Shazeer ¶ 29 (“The embedding layer 120 is configured to, for each network input in the input sequence, map the network input to a numeric representation of the network input in an embedding space, e.g., into a vector in the embedding space. The embedding layer 120 then provides the numeric representations of the network inputs to the first subnetwork in the sequence of encoder subnetworks 130, i.e., to the first encoder subnetwork 130 of the N encoder subnetworks 130.”))
in response to the encoding, generating an attention vector for each word of the one or more first natural language characters by determining how relevant, via weighting, each word is relative to at least one other word and relative to each word in the second partial string, and wherein the generating of the second partial string is based on the generation of an attention vector for each word (but see Shazeer ¶ 30 (“In particular, in some implementations, the embedding layer 120 is configured to map each network input to an embedded representation of the network input and then combine, e.g., sum or average, the embedded representation of the network input with a positional embedding of the input position of the network input in the input order to generate a combined embedded representation of the network input. That is, each position in the input sequence has a corresponding embedding and for each network input the embedding layer 120 combines the embedded representation of the network input with the embedding of the network input's position in the input sequence. Such positional embeddings can enable the model to make full use of the order of the input sequence without relying on recurrence or convolutions.”)).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified Walters to incorporate the teachings of Shazeer to encode and embed the words and other information to generate synthetic data, at least because doing so would enable creating content for multiple different contexts associated with different syntax generators. See Walters ¶ 35.
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to SHAHID KHAN whose telephone number is (571)270-0419. The examiner can normally be reached M-F, 9-5 est.
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, Usmaan Saeed can be reached at (571)272-4046. 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.
/SHAHID K KHAN/Primary Examiner, Art Unit 2146
Prosecution Insights