DETAILED ACTION
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
In response to Applicant’s claims filed on November 13, 2025, claims 1-11 are now pending for examination in the application.
Response to Arguments
This office action is in response to amendment filed 11/13/2025. In this action claim(s) 1-3, 5-11 is/are rejected under 35 U.S.C. 103 as being unpatentable over Tran (US Pub. No. 20220237368) in view of Brecque (US Pub. No. 20240289642). The Brecque reference has been added to address the amendment of wherein the update agent comprises reinforcement learning employing symbolic reasoning to detect anomalous data elements from non-validated responses and user feedback, and automatically modify the first database to correct the anomalous data elements, thereby improving data retrieval speed and enabling an upgrade of the computer system without modification of its source code.
Applicant’s arguments:
In regards to claim 1 on Pages 11, applicant argues “For example, the claimed invention is a practical application of AI technology that improves computer functionality and provides a specific technical solution, that goes beyond an abstract idea. More specifically, for example, the claimed invention presents a specific computer-implemented method and system,” as recited in claim 1.
Examiner’s Reply:
Answering queries is not a technological improvement. The claims merely use a conversational agent associated with processed documents in order to answer questions. The determination, vectorizing, and identification of data in documents is a computer-implemented abstract mental process.
Applicant’s arguments:
In regards to claim 1 on Pages 12, applicant argues “Applicant respectfully notes, for example, wherein even if the claims were viewed as reciting an abstract concept, the ordered combination of reinforcement learning, symbolic reasoning, and dynamic self-updating of the vectorized databases provides significantly more than the alleged abstract idea under Step 2B, because the claims define a non-conventional, non-generic arrangement of computer components that yields measurable performance benefits. The amended claims therefore satisfy the requirements of §101 and should be withdrawn from rejection.,” as recited in claim 1.
Examiner’s Reply:
Question answering (eg conversational question answering using extracted information) is well-understood, routine, and conventional. The additional elements merely allow a user to update and validate the new responses given a certain amount of processing resources.
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.
Claims 1‐11 are 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 pre‐AIA the applicant regards as the invention.
Claims 1, 10, and 11 recite(s) the limitation “hereby improving data retrieval speed and enabling an upgrade of the computer system without modification of its source code". This is an intended result.
Claims 2‐9 are rejected as dependent on claim(s) 1, 10, and 11.
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 1-11 is/are rejected under 35 U.S.C. 101 because the claimed invention is directed to non-patentable subject matter. The claims are directed to an abstract idea without significantly more.
Claim 1-11 is/are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more. The judicial exception is not integrated into a practical application. The claims do not include additional elements that are sufficient to amount to significantly more than judicial exception. The eligibility analysis in support of these findings is provided below, on Claim Rejections - 35 USC 101 accordance with the "2019 Revised Patent Subject Matter Eligibility Guidance" (published on 1/7/2019 in Fed, Register, Vol. 84, No. 4 at pgs. 50-57, hereinafter referred to as the "2019 PEG").
Step 1. in accordance with Step 1 of the eligibility inquiry (as explained in MPEP 2106), it is first noted the method (claims 1-9), system (claim 10), program (claim 11) are directed to one of the eligible categories of subject matter and therefore satisfies Step 1.
Step 2A. In accordance with Step 2A, prong one of the 2019 PEG, it is noted that the independent claims recite an abstract idea falling within the Mental Processes & Mathematical Concepts enumerated groupings of abstract ideas set forth in the 2019 PEG. Examiner is of the position that independent claims 1, 10, and 11 are directed towards the Mental Process Grouping of Abstract Ideas.
Independent claim 1, 10, 11 recites the following limitations directed towards a Mental Processes & Mathematical Concepts:
converting the electronic documents in the first database into vectors of numbers (The limitation recites a mathematical concept of converting data into vectors);
inserting said vectors into the second vector database (The limitation recites a mental process of observation and/or evaluation capable of being performed by the human mind by using computer as a tool to insert vectors);
generatingobservation and/or evaluation capable of being performed by the human mind by using computer as a tool to generate a response);
automatically identifying, by an update agent of the computer system, a source electronic document to be updated in the first database based on one or more of the query and the context data (The limitation recites a mental process of observation and/or evaluation capable of being performed by the human mind by using computer as a tool to identify a document);
automatically determining, by the update agent interacting with the generative artificial intelligence agent, an update action for the source electronic document that is identified based on one or more of the context data and the query
wherein the generative artificial intelligence agent
automatically executing by the update agent an update of the first database based on the update action that is determined (The limitation recites a mental process of observation and/or evaluation capable of being performed by the human mind by using computer as a tool to updating a database)
then converting the electronic documents in the first database into vectors of numbers once again to update the second vector database (The limitation recites a mental process of observation and/or evaluation capable of being performed by the human mind by using computer as a tool to convert data into vectors);
wherein the update agent comprises reinforcement learning employing symbolic reasoning to detect anomalous data elements from non-validated responses and user feedback (The limitation recites a mental process of observation and/or evaluation capable of being performed by the human mind by using computer as a tool to detect data elements), and
automatically modify the first database to correct the anomalous data elements, thereby improving data retrieval speed and enabling an upgrade of the computer system without modification of its source code (The limitation recites a mental process of observation and/or evaluation capable of being performed by the human mind by using computer as a tool to modify data elements).
Step 2A. In accordance with Step 2A, prong two of the 2019 PEG, the judicial exception is not integrated into a practical application because of the recitation in claim(s) 1, 10, and 11:
a first database that stores electronic documents (i.e., as a generic processor/component performing a generic computer function);
a second vector database (i.e., as a generic processor/component performing a generic computer function);
a processor (i.e., as a generic processor/component performing a generic computer function)on which a conversational agent and a generative artificial intelligence agent are installed, and wherein said processor comprises instructions configured to implement a computer-implemented method of managing the electronic documents stored in said first database, said computer-implemented method comprising
receiving, via the conversational agent of the computer system, a query to search for information in the electronic documents (recites insignificant extra solution activity that amounts to mere data gathering);
in an event of non-validation of the response, receiving, via the conversational agent, context data related to circumstances of said non-validation of the response (recites insignificant extra solution activity that amounts to mere data gathering).
Step 2B. Similar to the analysis under 2A Prong Two, the claim(s) does/do not include additional elements that are sufficient to amount to significantly more than the judicial exception. Because the additional elements of the independent claims amount to insignificant extra solution activity and/or mere instructions, the additional elements do not add significantly more to the judicial exception such that the independent claims as a whole would be patent eligible.
Therefore, independent claim(s) 1, 10, and 11 is/are rejected under 35 U.S.C. 101.
With respect to claim(s) 2:
Step 2A Prong One Analysis:
wherein the converting the electronic documents in the first database into vectors, and the inserting said vectors into the second vector database are executed once again after updating the first database (The limitation recites a mathematical concept of converting data into vectors).
Step 2A Prong Two Analysis:
This judicial exception is not integrated into a practical application because there are no
additional elements to provide practical application.
Step 2B Analysis:
The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception. The claim is not patent eligible.
With respect to claim(s) 3:
wherein the executing the update of the first database is executed (The limitation recites a mental process of evaluation and/or judgement capable of being performed by the human mind by executing an update)
after determining a plurality of update actions from a plurality of non-validated responses (The limitation recites a mental process of evaluation and/or judgement capable of being performed by the human mind by determining an action).
Step 2A Prong Two Analysis:
This judicial exception is not integrated into a practical application because there are no
additional elements to provide practical application.
Step 2B Analysis:
The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception. The claim is not patent eligible.
With respect to claim(s) 4:
wherein the executing the update of the first database is executed (The limitation recites a mental process of evaluation and/or judgement capable of being performed by the human mind by executing an update)
if a number of non-validated responses of said plurality of non-validated responses has reached a predefined threshold (The limitation recites a mental process of evaluation and/or judgement capable of being performed by the human mind by determining a threshold).
Step 2A Prong Two Analysis:
This judicial exception is not integrated into a practical application because there are no
additional elements to provide practical application.
Step 2B Analysis:
The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception. The claim is not patent eligible.
With respect to claim(s) 5:
describing the electronic documents of the first database and operational parameters for controlling operations, implemented in order to execute at least one of the converting the electronic documents into said vectors,
the generating the response to the query (The limitation recites a mental process of evaluation and/or judgement capable of being performed by the human mind by generating a response), and
the executing the update of the second vector database (The limitation recites a mental process of evaluation and/or judgement capable of being performed by the human mind by updating a database).
Step 2A Prong Two Analysis:
further comprising reading a governance computer file, stored in memory (recites insignificant extra solution activity that amounts to reading data).
Step 2B Analysis:
The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception. The claim is not patent eligible.
With respect to claim(s) 6:
generating metadata associated with each data vector, describing information about the source electronic document from which said each data vector originates (The limitation recites a mental process of evaluation and/or judgement capable of being performed by the human mind by generating metadata),
said metadata being inserted into the second vector database in association with a corresponding data vector (The limitation recites a mental process of evaluation and/or judgement capable of being performed by the human mind by inserting metadata).
Step 2A Prong Two Analysis:
This judicial exception is not integrated into a practical application because there are no additional elements to provide practical application.
Step 2B Analysis:
The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception. The claim is not patent eligible.
With respect to claim(s) 7:
wherein the generating the response to the query (The limitation recites a mental process of evaluation and/or judgement capable of being performed by the human mind by generating a response)
comprises extracting at least one keyword from the query (The limitation recites a mental process of evaluation and/or judgement capable of being performed by the human mind by extracting a keyword) and
comparing the at least one keyword that is extracted and the metadata in the second vector database (The limitation recites a mental process of evaluation and/or judgement capable of being performed by the human mind by comparing a keyword).
Step 2A Prong Two Analysis:
This judicial exception is not integrated into a practical application because there are no additional elements to provide practical application.
Step 2B Analysis:
The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception. The claim is not patent eligible.
With respect to claim(s) 8:
wherein the converting the electronic documents into said vectors comprises
splitting each electronic document of said electronic documents into smaller objects (The limitation recites a mental process of evaluation and/or judgement capable of being performed by the human mind by splitting a document);
converting each object of said smaller objects into a vector of numbers in a multi-dimensional space (The limitation recites a mathematical concept of converting data into vectors);
indexing the vectors (The limitation recites a mental process of evaluation and/or judgement capable of being performed by the human mind by indexing vectors).
Step 2A Prong Two Analysis:
This judicial exception is not integrated into a practical application because there are no additional elements to provide practical application.
Step 2B Analysis:
The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception. The claim is not patent eligible.
With respect to claim(s) 9:
wherein the converting the electronic documents into said vectors (The limitation recites a mathematical concept of converting data into vectors) comprises a preliminary step of pre-processing data of the electronic documents to one or more of correct and eliminate errors in the electronic documents (The limitation recites a mental process of evaluation and/or judgement capable of being performed by the human mind by preprocessing data).
Step 2A Prong Two Analysis:
This judicial exception is not integrated into a practical application because there are no additional elements to provide practical application.
Step 2B Analysis:
The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception. The claim is not patent eligible.
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.
Claim(s) 1-3, 5-11 is/are rejected under 35 U.S.C. 103 as being unpatentable over Tran (US Pub. No. 20220237368) in view of Brecque (US Pub. No. 20240289642).
With respect to claim 1, Tran teaches a computer-implemented method of managing electronic documents stored in a first database, said computer-implemented method implemented by a computer system and comprising steps, performed by the computer system, of:
converting the electronic documents in the first database into vectors of numbers (Paragraph 70 discloses encoder part is a GRU-RNN which generate a fixed length vector);
inserting said vectors into a second vector database (Paragraph 209-211 discloses Select deep neural network architecture (for example, retrieval, generative, and retrieve/refine, transformer-based, BERT-based, GPT-based, among others) for a learning machine Collect training data and update on periodic basis: Store non-public information into a database from a site desiring to have a chatbot to answer questions);
receiving, via a conversational agent of the computer system, a query to search for information in the electronic documents (Paragraph 211-212 discloses Store non-public information into a database from a site desiring to have a chatbot to answer questions, including CRM databases for common user questions and non-public product maintenance or service information for products Crawl web site of the company desiring to have the chatbot to answer questions to extract user manuals, FAQs and all publicly available text);
generating, via a generative artificial intelligence agent of the computer system using the second vector database, a response to the query (Paragraph(S) 130-132 discloses generate targeted responses/proposals for the user. The process includes: Select deep neural network architecture (for example, retrieval, generative, and retrieve/refine, transformer-based, BERT-based, GPT-based, among others) for a learning machine Train the learning machine with data that is logically grouped or clustered to provide context and accuracy (for example, by technology field or by industry/specialization);
in an event of non-validation of the response, receiving, via the conversational agent, context data related to circumstances of said non-validation of the response (Paragraph 240-243 discloses generate context-sensitive chats that are optimized to answer or interact with the user & If user is dissatisfied based on detected emotion, select another call-center agent best matched to the user profile or need and transfer to new selected agent or supervisor). Tran does not disclose identifying a source electronic document to be updated in the first database based on one or more of the query and the context data.
However, Brecque teaches automatically identifying, by an update agent of the computer system, a source electronic document to be updated in the first database based on one or more of the query and the context data (Paragraph 92 discloses the document generation system 300 may, in a case where the proposed change is to the one or more first terms, transmit the proposed change to the one or more first terms to the first source 302a and receive updated values for the one or more first terms from the first source 302a);
automatically determining, by the update agent interacting with the generative artificial intelligence agent, an update action for the source electronic document that is identified based on one or more of the context data and the query, wherein the generative artificial intelligence agent determines said update action for the source electronic document that is identified using the second vector database (Paragraph 95 discloses an AI controller 315 may apply symbolic reasoning to automatically trigger alerts related to the contract's lifecycle and to detect compatibility breaches between existing and new contracts at the level of the acceptable terms collector 305 before they are fully created and negotiated);
automatically executing, by an update agent, an update of the first database based on the update action that is determined, then converting the electronic documents in the first database into vectors of numbers once again to update the second vector database (Paragraph 176 discloses a TNN model for automatically extracting the structure of documents, to be encoded as a knowledge graph, is trained using large amounts of documents, while optimizing the TNNs parameters to perform the specific tasks of generating the structure (concepts/terms) for the documents and extracting the values associated with the concepts/terms in each document. In some embodiments of the invention, the input documents (contracts) are preprocessed, tokenized into words or sub-words (representing headings, sub-headings and terms, for example), and converted into numerical representations using word embeddings. Word embeddings are dense vectors that capture semantic relationships between words),
wherein the update agent comprises reinforcement learning employing symbolic reasoning to detect anomalous data elements from non-validated responses and user feedback, and automatically modify the first database to correct the anomalous data elements, thereby improving data retrieval speed and enabling an upgrade of the computer system without modification of its source code (Paragraph 92 discloses The AI issues scanner and fixer 310 may use symbolic reasoning to suggest to one or more parties new compatible terms which would resolve the issue and Paragraph 147 discloses rules to determine if users can or should work with each other can also be seamlessly changed and evolved in time based on user feedback and/or the quality of the recommendations).
Therefore, it would have been obvious before the effective filing data of invention was made to a person having ordinary skill in the art to modify Tran with Brecque. This would have facilitated improved conversational question answering using documents. See Brecque Paragraph(s) 3-26.
The Tran reference as modified by Brecque teaches all the limitations of claim 1. With respect to claim 2, Brecque teaches the computer-implemented method according to claim 1, wherein the converting the electronic documents in the first database into vectors, and the inserting said vectors into the second vector database are executed once again after updating the first database (Paragraph 173 discloses a TNN model for automatically extracting the structure of documents, to be encoded as a knowledge graph, is trained using large amounts of documents, while optimizing the TNNs parameters to perform the specific tasks of generating the structure (concepts/terms) for the documents and extracting the values associated with the concepts/terms in each document. In some embodiments of the invention, the input documents (contracts) are preprocessed, tokenized into words or sub-words (representing headings, sub-headings and terms, for example), and converted into numerical representations using word embeddings. Word embeddings are dense vectors that capture semantic relationships between words). The motivation to combine statement previously provided in the rejection of independent claim 1 provided above, combining the Tran reference and the Brecque reference is applicable to dependent claim 2.
The Tran reference as modified by Brecque teaches all the limitations of claim 1. With respect to claim 3, Brecque teaches the computer-implemented method according to claim 1, wherein the executing the update of the first database is executed after determining a plurality of update actions from a plurality of non-validated responses (Paragraph 120 discloses In some embodiments of the invention, the contract rules 432 may include logical rules 831 which prevent logically invalid combinations of terms in the contract terms 438, such as for example a notice period longer than the contract duration. The contract rules 432 may also include legal rules 832 that prevent illegal combinations of contract terms 309 from occurring and contract type specific rules 833 that verify that the combinations of contract terms 309 comply with acceptable terms 305 for that contract type 303). The motivation to combine statement previously provided in the rejection of independent claim 1 provided above, combining the Tran reference and the Brecque reference is applicable to dependent claim 3.
The Tran reference as modified by Brecque teaches all the limitations of claim 1. With respect to claim 5, Brecque teaches the computer-implemented method according to claim 1, further comprising reading a governance computer file, stored in memory, describing the electronic documents of the first database and operational parameters for controlling operations, implemented in order to execute at least one of the converting the electronic documents into said vectors, the generating the response to the query, and the executing the update of the second vector database (Paragraph 89 discloses the type of the document 303 defining a plurality of rules 432 governing terms of the document). The motivation to combine statement previously provided in the rejection of independent claim 1 provided above, combining the Tran reference and the Brecque reference is applicable to dependent claim 5.
The Tran reference as modified by Brecque teaches all the limitations of claim 1. With respect to claim 6, Tran teaches the computer-implemented method according to claim 1, further comprising generating metadata associated with each data vector, describing information about the source electronic document from which said each data vector originates, said metadata being inserted into the second vector database in association with a corresponding data vector (Paragraph 303 discloses capture information from the one or more entities, normalize the captured information from first and second manufacturers in a common format, and add metadata for the captured information).
The Tran reference as modified by Brecque teaches all the limitations of claim 6. With respect to claim 7, Tran teaches the computer-implemented method according to claim 6, wherein the generating the response to the query comprises extracting at least one keyword from the query and comparing the at least one keyword that is extracted and the metadata in the second vector database (Paragraph 99 discloses query-based summarization produces a summary that contains information which answers the input question).
The Tran reference as modified by Brecque teaches all the limitations of claim 1. With respect to claim 8, Tran teaches the computer-implemented method according to claim 1, wherein the converting the electronic documents into said vectors comprises
splitting each electronic document of said electronic documents into smaller objects (Paragraph 165-169 discloses generates context-sensitive text by: using a first learning machine to map text matching each topic to a corresponding vector; building a search index for the search topics and in response to a search topic returning a responsive first vector; at run time, using a second learning machine to map a topic to a second vector; determining similarity between the responsive first vector and the second vector, selecting the most responsive first vector and retrieving text for the most responsive first vector);
converting each object of said smaller objects into a vector of numbers in a multi-dimensional space (Paragraph 165-169 discloses generates context-sensitive text by: using a first learning machine to map text matching each topic to a corresponding vector; building a search index for the search topics and in response to a search topic returning a responsive first vector; at run time, using a second learning machine to map a topic to a second vector; determining similarity between the responsive first vector and the second vector, selecting the most responsive first vector and retrieving text for the most responsive first vector);
indexing the vectors (Paragraph 165-169 discloses generates context-sensitive text by: using a first learning machine to map text matching each topic to a corresponding vector; building a search index for the search topics and in response to a search topic returning a responsive first vector; at run time, using a second learning machine to map a topic to a second vector; determining similarity between the responsive first vector and the second vector, selecting the most responsive first vector and retrieving text for the most responsive first vector).
The Tran reference as modified by Brecque teaches all the limitations of claim 8. With respect to claim 9, Tran teaches the computer-implemented method according to claim 8, wherein the converting the electronic documents into said vectors comprises a preliminary step of pre-processing data of the electronic documents to one or more of correct and eliminate errors in the electronic documents (Paragraph 165-169 discloses generates context-sensitive text by: using a first learning machine to map text matching each topic to a corresponding vector; building a search index for the search topics and in response to a search topic returning a responsive first vector; at run time, using a second learning machine to map a topic to a second vector; determining similarity between the responsive first vector and the second vector, selecting the most responsive first vector and retrieving text for the most responsive first vector)..
With respect to claim 10, Tran teaches a computerized electronic document management system comprising:
a first database that stores electronic documents (Paragraph 164 discloses a database of patents can be searched to locate documents matching the text input, and then the matching documents);
a second vector database (Paragraph 209-211 discloses a database from a site desiring to have a chatbot to answer questions);
a processor on which a conversational agent, a generative artificial intelligence agent, and an update agent are installed, and wherein said processor comprises instructions configured to implement a computer-implemented method of managing the electronic documents stored in said first database, said computer-implemented method comprising
converting the electronic documents in the first database into vectors of numbers (Paragraph 70 discloses encoder part is a GRU-RNN which generate a fixed length vector);
inserting said vectors into a second vector database (Paragraph 209-211 discloses Select deep neural network architecture (for example, retrieval, generative, and retrieve/refine, transformer-based, BERT-based, GPT-based, among others) for a learning machine Collect training data and update on periodic basis: Store non-public information into a database from a site desiring to have a chatbot to answer questions);
receiving, via a conversational agent, a query to search for information in the electronic documents (Paragraph 211-212 discloses Store non-public information into a database from a site desiring to have a chatbot to answer questions, including CRM databases for common user questions and non-public product maintenance or service information for products Crawl web site of the company desiring to have the chatbot to answer questions to extract user manuals, FAQs and all publicly available text);
generating, via a generative artificial intelligence agent using the second vector database, a response to the query (Paragraph(S) 130-132 discloses generate targeted responses/proposals for the user. The process includes: Select deep neural network architecture (for example, retrieval, generative, and retrieve/refine, transformer-based, BERT-based, GPT-based, among others) for a learning machine Train the learning machine with data that is logically grouped or clustered to provide context and accuracy (for example, by technology field or by industry/specialization);
in an event of non-validation of the response, receiving, via the conversational agent, context data related to circumstances of said non-validation of the response (Paragraph 240-243 discloses generate context-sensitive chats that are optimized to answer or interact with the user & If user is dissatisfied based on detected emotion, select another call-center agent best matched to the user profile or need and transfer to new selected agent or supervisor). Tran does not disclose identifying a source electronic document to be updated in the first database based on one or more of the query and the context data.
However, Brecque teaches automatically identifying, by an update agent, a source electronic document to be updated in the first database based on one or more of the query and the context data (Paragraph 92 discloses the document generation system 300 may, in a case where the proposed change is to the one or more first terms, transmit the proposed change to the one or more first terms to the first source 302a and receive updated values for the one or more first terms from the first source 302a);
automatically determining, by the update agent interacting with the generative artificial intelligence agent, an update action for the source electronic document that is identified based on one or more of the context data and the query, wherein the generative artificial intelligence agent determines said update action for the source electronic document that is identified using the second vector database (Paragraph 95 discloses an AI controller 315 may apply symbolic reasoning to automatically trigger alerts related to the contract's lifecycle and to detect compatibility breaches between existing and new contracts at the level of the acceptable terms collector 305 before they are fully created and negotiated);
automatically executing, by an update agent, an update of the first database based on the update action that is determined, then converting the electronic documents in the first database into vectors of numbers once again to update the second vector database (Paragraph 176 discloses a TNN model for automatically extracting the structure of documents, to be encoded as a knowledge graph, is trained using large amounts of documents, while optimizing the TNNs parameters to perform the specific tasks of generating the structure (concepts/terms) for the documents and extracting the values associated with the concepts/terms in each document. In some embodiments of the invention, the input documents (contracts) are preprocessed, tokenized into words or sub-words (representing headings, sub-headings and terms, for example), and converted into numerical representations using word embeddings. Word embeddings are dense vectors that capture semantic relationships between words),
wherein the update agent comprises reinforcement learning employing symbolic reasoning to detect anomalous data elements from non-validated responses and user feedback, and automatically modify the first database to correct the anomalous data elements, thereby improving data retrieval speed and enabling an upgrade of the computer system without modification of its source code (Paragraph 92 discloses The AI issues scanner and fixer 310 may use symbolic reasoning to suggest to one or more parties new compatible terms which would resolve the issue and Paragraph 147 discloses rules to determine if users can or should work with each other can also be seamlessly changed and evolved in time based on user feedback and/or the quality of the recommendations).
Therefore, it would have been obvious before the effective filing data of invention was made to a person having ordinary skill in the art to modify Tran with Brecque. This would have facilitated improved conversational question answering using documents. See Brecque Paragraph(s) 3-26.
With respect to claim 11, Tran teaches a non-transitory computer program comprising instructions which, when they are executed by a processor, cause the processor to implement a computer-implemented method of managing electronic documents stored in a first database, said computer-implemented method comprising
converting the electronic documents in the first database into vectors of numbers (Paragraph 70 discloses encoder part is a GRU-RNN which generate a fixed length vector);
inserting said vectors into a second vector database (Paragraph 209-211 discloses Select deep neural network architecture (for example, retrieval, generative, and retrieve/refine, transformer-based, BERT-based, GPT-based, among others) for a learning machine Collect training data and update on periodic basis: Store non-public information into a database from a site desiring to have a chatbot to answer questions);
receiving, via a conversational agent installed on the processor, a query to search for information in the electronic documents (Paragraph 211-212 discloses Store non-public information into a database from a site desiring to have a chatbot to answer questions, including CRM databases for common user questions and non-public product maintenance or service information for products Crawl web site of the company desiring to have the chatbot to answer questions to extract user manuals, FAQs and all publicly available text);
generating, via a generative artificial intelligence agent installed on the processor, a response to the query, said generative artificial intelligence agent using the second vector database (Paragraph(S) 130-132 discloses generate targeted responses/proposals for the user. The process includes: Select deep neural network architecture (for example, retrieval, generative, and retrieve/refine, transformer-based, BERT-based, GPT-based, among others) for a learning machine Train the learning machine with data that is logically grouped or clustered to provide context and accuracy (for example, by technology field or by industry/specialization);
in an event of non-validation of the response, receiving, via the conversational agent, context data related to circumstances of said non-validation of the response (Paragraph 240-243 discloses generate context-sensitive chats that are optimized to answer or interact with the user & If user is dissatisfied based on detected emotion, select another call-center agent best matched to the user profile or need and transfer to new selected agent or supervisor). Tran does not disclose identifying a source electronic document to be updated in the first database based on one or more of the query and the context data.
However, Brecque teaches automatically identifying, by an update agent installed on the processor, a source electronic document to be updated in the first database based on one or more of the query and the context data (Paragraph 92 discloses the document generation system 300 may, in a case where the proposed change is to the one or more first terms, transmit the proposed change to the one or more first terms to the first source 302a and receive updated values for the one or more first terms from the first source 302a);
automatically determining, by the update agent interacting with the generative artificial intelligence agent, an update action for the source electronic document that is identified based on one or more of the context data and the query, wherein the generative artificial intelligence agent determines said update action for the source electronic document that is identified using the second vector database (Paragraph 95 discloses an AI controller 315 may apply symbolic reasoning to automatically trigger alerts related to the contract's lifecycle and to detect compatibility breaches between existing and new contracts at the level of the acceptable terms collector 305 before they are fully created and negotiated);
automatically executing, by an update agent, an update of the first database based on the update action that is determined, then converting the electronic documents in the first database into vectors of numbers once again to update the second vector database (Paragraph 176 discloses a TNN model for automatically extracting the structure of documents, to be encoded as a knowledge graph, is trained using large amounts of documents, while optimizing the TNNs parameters to perform the specific tasks of generating the structure (concepts/terms) for the documents and extracting the values associated with the concepts/terms in each document. In some embodiments of the invention, the input documents (contracts) are preprocessed, tokenized into words or sub-words (representing headings, sub-headings and terms, for example), and converted into numerical representations using word embeddings. Word embeddings are dense vectors that capture semantic relationships between words),
wherein the update agent comprises reinforcement learning employing symbolic reasoning to detect anomalous data elements from non-validated responses and user feedback, and automatically modify the first database to correct the anomalous data elements, thereby improving data retrieval speed and enabling an upgrade of the computer system without modification of its source code (Paragraph 92 discloses The AI issues scanner and fixer 310 may use symbolic reasoning to suggest to one or more parties new compatible terms which would resolve the issue and Paragraph 147 discloses rules to determine if users can or should work with each other can also be seamlessly changed and evolved in time based on user feedback and/or the quality of the recommendations).
Therefore, it would have been obvious before the effective filing data of invention was made to a person having ordinary skill in the art to modify Tran with Brecque. This would have facilitated improved conversational question answering using documents. See Brecque Paragraph(s) 3-26.
Claim(s) 4 is/are rejected under 35 U.S.C. 103 as being unpatentable over Tran (US Pub. No. 20220237368) and Brecque (US Pub. No. 20240289642) in view of Mankovskii (US Pub. No. 20200134103).
The Tran reference as modified by Brecque teaches all the limitations of claim 1. With respect to claim 4, Tran as modified by Brecque does not disclose a predefined threshold.
However, teaches the computer-implemented method according to claim 3, wherein the executing the update of the first database is executed if a number of non-validated responses of said plurality of non-validated responses has reached a predefined threshold (Paragraph 59 discloses incrementally adjust criteria responsive to a given instance of feedback, for instance, feedback indicating that a caption is not relevant or is not descriptive may cause criteria to be adjusted in a way that makes the criteria less likely to be satisfied, for instance raising a threshold or lowering threshold. In some embodiments, the criteria of narrative text descriptions that are not presented may be adjusted based on feedback, for instance those narrative text descriptions adjacent a threshold that were filtered out for relevance may be adjusted responsive to feedback indicating that those that were not filtered out were not relevant or descriptive).
Therefore, it would have been obvious before the effective filing data of invention was made to a person having ordinary skill in the art to modify Tran and Brecque with Mankovskii. This would have facilitated improved conversational question answering using documents. See Mankovskii Paragraph(s) 5-8.
Relevant Prior Art
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure. The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
US PG-Pub. No. 20240370479 is directed to SEMANTIC SEARCH AND SUMMARIZATION FOR ELECTRONIC DOCUMENTS: [0054] sending a natural language generation (NLG) request to a generative artificial intelligence (AI) model. The generative AI model may comprise a machine learning model that implements a large language model (LLM) to support natural language processing (NLP) operations, such as natural language understanding (NLU), natural language generation (NLG), and other NLP operations. The NLG request may request an abstractive summary of document content for a subset of candidate document vectors from the set of candidate document vectors. The abstractive summary may comprise a natural language representation of the human language. The method may include receiving a NLG response with the abstractive summary from the generative AI model. Other embodiments are described and claimed.
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 NICHOLAS E ALLEN whose telephone number is (571)270-3562. The examiner can normally be reached Monday through Thursday 830-630.
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, Boris Gorney can be reached at (571) 270-5626. 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.
/N.E.A/Examiner, Art Unit 2154
/BORIS GORNEY/Supervisory Patent Examiner, Art Unit 2154