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 .
This action is in response to the communication filed on May 16, 2024.
Claims 1-20 are examined and are pending.
Claim Rejections - 35 USC § 103
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
Claims 1-2, 4, 6-18 are rejected under 35 U.S.C. 103 as being unpatentable over Wang (US 2023/0245651 A1), in view of Jones et al (US 2020/0410505 A1).
As per claim 1, Wang discloses:
- a computer-implemented method, the method comprising (Para [0005], line 1-5, “a method for enabling user-centered and contextually relevant conversational interaction”),
- obtaining, by a computing system comprising one or more computing devices, training data comprising (Para [0079], [0081], [0106], [0116] – [0117], [0188], Fig. 5, item 502, 503, training data and AI systems can use reasoning techniques to predict user intent associated with user input query, and Fig. 22, item 2204 – 2209, Para [0053], [0350], [0353], recommendation response associated with user interaction or user query),
- providing, by the computing system, a first portion of the training data to a sequence processing model as an input training example and receiving at least one output (Fig. 16, item 1604-1606, Fig. 20, item 2003, Para [0079], [0119], [0437], system provide collected data to ECAP model, CM and CAM model (i.e., sequence of processing model)
- comparing, by the computing system, the at least one output to a second portion of the training data to generate at least one evaluation component (Para [0275], validating outputs by comparing them with ground truth data or known outcome or other relevant metrics (i.e., comparing output)),
- and modifying, by the computing system, the sequence processing model based at least in part on the at least one evaluation component (Fig. 22, item 2207, Para [0081], line 45-50, Para [0120], [0351], [0358], updating models according to model performance or based on feedback and new data),
Wang does not explicitly disclose a plurality of triplets, each triplet comprising an example query. However, in the same field of endeavor Jones in an analogous art discloses a plurality of triplets, each triplet comprising an example query (Abstract, line 3-10, Para [0042], [0066], set of triplets (i.e., plurality of triplet) with question and answer (i.e., query), intent and sentiment (i.e., reasoning) and response).
Therefore, it would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to incorporate a plurality of triplet and each triplet comprising a query taught by Jones as the means to train a model for conversational recommendation in Wang, (Wang, Para [0079], Jones, Para [0005]). Wang and Jones are analogous prior art since they both deal with natural language conversation processing and model recommendation between a user and an agent. A person of the ordinary skill in the art would have been motivated to make aforementioned modification to effective communication between a user and an agent and determine user intent in a chat session. This is because one aspect of Wang invention is to provide improved accuracy and efficiency in understanding contextual information in the environment and predicting the most relevant user intent and objective as described at least in Para [0033]. A plurality of triplet where each triplet with an example query, reason and response are part of this model recommendation in a chat session. However, Wang doesn’t specify any particular manner in which a plurality of triplet where each triplet with an example query, reason and response are processed. This would have lead one of the ordinary skill in the art to seek and recognize the plurality of triplet where each triplet with an example query, reason and response as taught by Jones. Jpnes describes how their chat transcript system analyzes historical chat transcript data in order to generate and suggest a response to a customer representative who may be faced with a similar question as was included in the historical chat transcript data, in order to improve the effectiveness of automated customer representative tools, as described at least in Para [0030], as desired by Wang.
As per claim 2, rejection of claim 1 is incorporated, and further Wang discloses:
- providing, by the computing system, the first portion of the training data to the sequence processing model as an input training example and receiving at least one output, (Para [0129], [0134] – [0135], receiving user input and comparing results for desirable outcome (i.e., training data as input and receiving output)),
- providing the example query to the sequence processing model as a first training example input and receiving a first output including a generated model reasoning plan (Para [0081], [0491], [0250], [0190], providing user request or query (i.e., example query) to processing model and receiving output including reasoning plan),
- and providing the example query and the example model reasoning plan to the sequence processing model as a second training example input and receiving a second output including a generated model recommendation response (Fig. 22, item 2204 – 2209, Para [0053], [0350], [0353], recommendation response associated with user interaction or user query),
Wang does not explicitly disclose for each triplet of at least a subset of the triplets. However, in the same field of endeavor Jones in an analogous art discloses a plurality of for each triplet of at least a subset of the triplets (Abstract, line 3-10, Para [0042], [0066], set of triplets (i.e., plurality of triplet) with question and answer (i.e., query), intent and sentiment (i.e., reasoning) and response).
Therefore, it would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to incorporate a plurality of triplet and each triplet comprising a query taught by Jones as the means to train a model for conversational recommendation in Wang, (Wang, Para [0079], Jones, Para [0005]). Wang and Jones are analogous prior art since they both deal with natural language conversation processing and model recommendation between a user and an agent. A person of the ordinary skill in the art would have been motivated to make aforementioned modification to effective communication between a user and an agent and determine user intent in a chat session. This is because one aspect of Wang invention is to provide improved accuracy and efficiency in understanding contextual information in the environment and predicting the most relevant user intent and objective as described at least in Para [0033]. A plurality of triplet where each triplet with an example query, reason and response are part of this model recommendation in a chat session. However, Wang doesn’t specify any particular manner in which a plurality of triplet where each triplet with an example query, reason and response are processed. This would have lead one of the ordinary skill in the art to seek and recognize the plurality of triplet where each triplet with an example query, reason and response as taught by Jones. Jpnes describes how their chat transcript system analyzes historical chat transcript data in order to generate and suggest a response to a customer representative who may be faced with a similar question as was included in the historical chat transcript data, in order to improve the effectiveness of automated customer representative tools, as described at least in Para [0030], as desired by Wang.
As per claim 4, rejection of claim 2 is incorporated, and further Wang discloses:
- providing, by the computing system, at least a portion of the training data to the sequence processing model and receiving at least one response, comprises, (Para [0129], [0134] – [0135], receiving user input and comparing results for desirable outcome (i.e., training data as input and receiving output)),
providing the example query to the sequence processing model as a first training example input and receiving a third output including a generated model reasoning plan (Para [0081], [0491], [0250], [0190], providing user request or query (i.e., example query) to processing model and receiving output including reasoning plan),
- wherein: the subset is a first subset of the plurality of triplets (Para [0042], [0066], set of triplets (i.e., plurality of triplet) with question and answer (i.e., query), intent and sentiment (i.e., reasoning) and response).
Wang does not explicitly disclose for each triplet of a second subset of the plurality of triplets. However, in the same field of endeavor Jones in an analogous art discloses for each triplet of a second subset of the plurality of triplets (Abstract, line 3-10, Para [0042], [0066], set of triplets (i.e., plurality of triplet) with question and answer (i.e., query), intent and sentiment (i.e., reasoning) and response).
Therefore, it would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to incorporate a plurality of triplet and each triplet comprising a query taught by Jones as the means to train a model for conversational recommendation in Wang, (Wang, Para [0079], Jones, Para [0005]). Wang and Jones are analogous prior art since they both deal with natural language conversation processing and model recommendation between a user and an agent. A person of the ordinary skill in the art would have been motivated to make aforementioned modification to effective communication between a user and an agent and determine user intent in a chat session. This is because one aspect of Wang invention is to provide improved accuracy and efficiency in understanding contextual information in the environment and predicting the most relevant user intent and objective as described at least in Para [0033]. A plurality of triplet where each triplet with an example query, reason and response are part of this model recommendation in a chat session. However, Wang doesn’t specify any particular manner in which a plurality of triplet where each triplet with an example query, reason and response are processed. This would have lead one of the ordinary skill in the art to seek and recognize the plurality of triplet where each triplet with an example query, reason and response as taught by Jones. Jones describes how their chat transcript system analyzes historical chat transcript data in order to generate and suggest a response to a customer representative who may be faced with a similar question as was included in the historical chat transcript data, in order to improve the effectiveness of automated customer representative tools, as described at least in Para [0030], as desired by Wang.
As per claim 6, rejection of claim 1 is incorporated, and further Wang discloses:
- wherein: the sequence processing model is a large language model (Para [0136], model is a language model (i.e., large language model).
As per claim 7, rejection of claim 1 is incorporated, and further Wang discloses:
- obtaining, by the computing system, a user query for the sequence processing model (Para [0154], Fig. 24, item 2401, user input is a request or query (i.e., obtaining user query)),
- providing an input prompt to the sequence processing model, the input prompt including the user query, a conversation history associated with the user query, and a model preamble (Fig. 7, item 708, Para [0210], input prompt with user query and Para [0244], dialogue history maintains a record of the previous turn (i.e., conversation history).
As per claim 8, rejection of claim 7 is incorporated, and further Wang discloses:
- wherein: the model preamble includes instructions indicative of one or more external tools available to the sequence processing model (Para [0215], communicating with external system to receive relevant data (i.e., instruction indicating external tool)).
As per claim 9, rejection of claim 1 is incorporated, and further Wang discloses:
- wherein: the sequence processing model is configured to read and write data to a conversational data store as part of processing a user query (Para [0165], converting spoken language into written text (i.e., read and write data)).
As per claim 10, rejection of claim 1 is incorporated, and further Wang discloses:
- wherein: the example model reasoning plan of the training data includes data indicative of at least one of: a goal associated with the example query; a situation associated with the example query; a consideration associated with the example query; and a plan associated with the example query (Para [0142], goal or intention of the query, Para [0299], understanding the current situation of the user environment).
As per claim 11, rejection of claim 1 is incorporated, and further Wang discloses:
- wherein: the example recommendation response of the training data is associated with at least one model interaction turn for generating at least one recommendation in response to the example query of the training data (Para [0222], [0244], interaction with previous turn in conversation recommendation).
As per claim 12, Wang discloses:
- a computing system, comprising (Para [0501], the present invention may be a system, a method, and/or a computer program product at any possible technical detail level of integration),
- one or more processors; one or more computer-readable storage media that collectively store a recommendation system, the recommendation system comprising (Para [0501], computer program product may include a computer readable storage medium (or media) having computer readable program instructions thereon for causing a processor to cany out aspects of the present invention),
- a conversational user interface configured to receive a user query and provide a recommendation response (Para [0036], [0156], [0157], conversational user interface for querying or searching and recommended response),
- a machine-learned sequence processing model that has been trained on training data including (Fig. 24, item 2402, Para [0081], [0120], (Para [0032], training machine learning model (ML), and (Fig. 16, item 1604-1606, Fig. 20, item 2003, Para [0079], [0119], [0437], system provide collected data to ECAP model, CM and CAM model (i.e., sequence of processing model),
- receive an input prompt including the user query, a previous conversation history associated with the user query, and a model preamble (Fig. 7, item 708, Para [0210], input prompt with user query and Para [0244], dialogue history maintains a record of the previous turn (i.e., conversation history),
- generate a reasoning plan for responding to the user query (Para [0150], [0229], planning for the user query),
- generate a model response based at least in part on the user query and the reasoning plan (Fig. 16, item 1606, Fig. 22, item 2209, Para [0027], Para [0088], [0091], generating model response according to the plan),
Wang does not explicitly disclose a plurality of triplets, each triplet comprising an example query. However, in the same field of endeavor Jones in an analogous art discloses a plurality of triplets, each triplet comprising an example query (Abstract, line 3-10, Para [0042], [0066], set of triplets (i.e., plurality of triplet) with question and answer (i.e., query), intent and sentiment (i.e., reasoning) and response).
Therefore, it would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to incorporate a plurality of triplet and each triplet comprising a query taught by Jones as the means to train a model for conversational recommendation in Wang, (Wang, Para [0079], Jones, Para [0005]). Wang and Jones are analogous prior art since they both deal with natural language conversation processing and model recommendation between a user and an agent. A person of the ordinary skill in the art would have been motivated to make aforementioned modification to effective communication between a user and an agent and determine user intent in a chat session. This is because one aspect of Wang invention is to provide improved accuracy and efficiency in understanding contextual information in the environment and predicting the most relevant user intent and objective as described at least in Para [0033]. A plurality of triplet where each triplet with an example query, reason and response are part of this model recommendation in a chat session. However, Wang doesn’t specify any particular manner in which a plurality of triplet where each triplet with an example query, reason and response are processed. This would have lead one of the ordinary skill in the art to seek and recognize the plurality of triplet where each triplet with an example query, reason and response as taught by Jones. Jpnes describes how their chat transcript system analyzes historical chat transcript data in order to generate and suggest a response to a customer representative who may be faced with a similar question as was included in the historical chat transcript data, in order to improve the effectiveness of automated customer representative tools, as described at least in Para [0030], as desired by Wang.
As per claim 13, rejection of claim 12 is incorporated, and further Wang discloses:
- wherein: the model response includes at least one of an indication of an objective from the user query, an indication of relevant and established facts associated with the user query, an indication of key consideration points, one or more recommendations and a justification for each recommendation, or one or more follow-up questions or invitations (Fig. 10, item 1010, Fig. 11, item 1104, Pata [0005], [0116]-[0117], intent and objective of the user query or interaction).
As per claim 14, rejection of claim 12 is incorporated, and further Wang discloses:
- wherein: the machine-learned sequence processing model is configured to, in response to the input prompt (Para [0197], [0210], response to the user prompt),
- generate computer-executable code to access one or more external computing services using one or more application programming interfaces (Para [0252], [0335], code for external services).
As per claim 15, rejection of claim 14 is incorporated, and further Wang discloses:
- wherein: the reasoning plan is based at least in part on the computer-executable code (Para [0038], plan based on software compiled code, microcode and the like).
As per claim 16, rejection of claim 14 is incorporated, and further Wang discloses:
- wherein: the model response is based at least in part on the computer-executable code (Para [0227], [0252], response based on code).
As per claim 17, rejection of claim 12 is incorporated, and further Wang discloses:
- wherein: the machine-learned sequence processing model is configured to, in response to the input prompt (Para [0197], [0210], response to the user prompt),
- request, from one or more memories, data associated with one or more previous conversational turns associated with the user query (Fig. 15, item 1501, Para [0244], [0412], storing interaction history (i.e., previous conversation),
- and store, in the one or more memories, data associated with a current conversational turn associated with the user query (Para [0348], [0453], storing data associated with current conversation)).
Claims 3, 5 and 19-20 are rejected under 35 U.S.C. 103 as being unpatentable over Wang (US 2023/0245651 A1), in view of Jones et al (US 2020/0410505 A1), as applied to claims 1, 12 and 18 and further in view of Wu et al (US 2022/0139384 A1).
As per claim 3, rejection of claim 2 is incorporated,
Combined method of Wang and Jones does not explicitly disclose evaluating a first loss component based on comparing the generated model reasoning plan to the example model reasoning plan from the triplet of the training data; and evaluating a second loss component based on comparing the generated model recommendation response to the example recommendation response from the triplet of the training data; wherein modifying, by the computing system, the sequence processing model comprises modifying the sequence processing model based at least in part on the first loss component and the second loss component. However, in the same field of endeavor Wu in an analogous art discloses evaluating a first loss component based on comparing the generated model reasoning plan to the example model reasoning plan from the triplet of the training data (Para [0020], [0021], calculating one or more loss function (i.e., evaluating first loss component),, and evaluating a second loss component based on comparing the generated model recommendation response to the example recommendation response from the triplet of the training data (Para [0020], [0021], calculating one or more loss function (i.e., evaluating second loss component),wherein modifying, by the computing system, the sequence processing model comprises modifying the sequence processing model based at least in part on the first loss component and the second loss component (Fig. 3, Para [0046], [0047], item 360, updating the model based on loss).
Therefore, it would have been obvious to a person of the ordinary skill in the art before the effective filing date of the claimed invention to combine the teaching of Wang, as previously modified with Jones, with the teaching of Wu by modifying Wang such that losses are calculated between different model in conversation recommendation modeling technique. The motivation for doing so would be to detect higher accuracy between different models, (Wu, Para [0051]).
As per claim 5, rejection of claim 4 is incorporated,
Combined method of Wang and Jones does not explicitly disclose wherein: comparing, by the computing system, the at least one output to the second portion of the training data to generate the at least one evaluation component, comprises; evaluating a third loss component based on comparing the example model reasoning plan to a generated model reasoning plan data; modifying, by the computing system, the sequence processing model comprises, for the second subset of the plurality of triplets, modifying the sequence processing model based on the third loss component without calculating a loss based on a generated model recommendation response. However, in the same field of endeavor Wu in an analogous art disclose wherein: comparing, by the computing system, the at least one output to the second portion of the training data to generate the at least one evaluation component, comprises (Para [0054], comparing the output between language model), evaluating a third loss component based on comparing the example model reasoning plan to a generated model reasoning plan data (Para [0020], [0021], calculating one or more loss function (i.e., evaluating first loss component), modifying, by the computing system, the sequence processing model comprises, for the second subset of the plurality of triplets, modifying the sequence processing model based on the third loss component without calculating a loss based on a generated model recommendation response (Fig. 3, Para [0046], [0047], item 360, updating the model based on loss).
Therefore, it would have been obvious to a person of the ordinary skill in the art before the effective filing date of the claimed invention to combine the teaching of Wang, as previously modified with Jones, with the teaching of Wu by modifying Wang such that losses are calculated between different model in conversation recommendation modeling technique. The motivation for doing so would be to detect higher accuracy between different models, (Wu, Para [0051]).
As per claims 18-20,
Claims 18-20 are computer readable medium claims corresponding to method claims 1, 2-3 and 4-5 respectfully and rejected under the same reason set forth to the rejection of claim 1, 2-3 and 4-5 above.
Contact Information
6. Any inquiry concerning this communication or earlier communications from the examiner should be directed to MOHAMMED R UDDIN whose telephone number is (571)270-3138. The examiner can normally be reached M-F: 9:00 AM-5:00 PM.
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, Apu Mofiz can be reached at (571) 272-4080. 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.
/MOHAMMED R UDDIN/Primary Examiner, Art Unit 2161
Prosecution Insights