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 .
Drawings
The drawings are objected to as failing to comply with 37 CFR 1.84(p)(5) because they include the following reference character not mentioned in the description:
“100” in Figure 1.
Corrected drawing sheets in compliance with 37 CFR 1.121(d), or amendment to the specification to add the reference character(s) in the description in compliance with 37 CFR 1.121(b) are required in reply to the Office action to avoid abandonment of the application. Any amended replacement drawing sheet should include all of the figures appearing on the immediate prior version of the sheet, even if only one figure is being amended. Each drawing sheet submitted after the filing date of an application must be labeled in the top margin as either “Replacement Sheet” or “New Sheet” pursuant to 37 CFR 1.121(d). If the changes are not accepted by the examiner, the applicant will be notified and informed of any required corrective action in the next Office action. The objection to the drawings will not be held in abeyance.
Specification
The disclosure is objected to because of the following informalities:
In paragraph 0031, line 6, “one or more a machine learning (ML) frameworks or models” should read “one or more machine learning (ML) frameworks or models”.
Appropriate correction is required.
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 15 – 20 are rejected under 35 U.S.C. 101 because the claimed invention is directed to non-statutory subject matter. The claims do not fall within at least one of the four categories of patent eligible subject matter because the broadest reasonable interpretation of a “computer-readable storage medium” can encompass non-statutory transitory forms of signal transmission, such as a propagating electrical or electromagnetic signal per se.
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, 3, 6 – 8, 10, 13 – 15, 17 and 20 are rejected under 35 U.S.C. 103 as being unpatentable over Taylert et al. (US Patent No. 11,960,514), hereinafter Taylert, in view of Penta et al. (US Patent No. 12,373,506), hereinafter Penta.
Regarding claim 1, Taylert discloses an apparatus comprising:
a memory; and a processor coupled to the memory (Column 22, lines 26-30, "A machine implementing the techniques herein comprises a hardware processor, and non-transitory computer memory holding computer program instructions that are executed by the processor to perform the above-described methods."), the processor configured to:
receive interaction content from an interaction session of a source device (Column 14, lines 23-40, "With the above as background, the following describes the interactive conversation (e.g., live-chat) assistance method and system of this disclosure. As a shorthand, and without intending to be limiting, this functionality is referred to as “suggested replies.” As will be seen, the approach leverages a generative AI (e.g., a language model such as OpenAI ChatGPT) to facilitate generation of a reply to an utterance that is received by the system during the interactive conversation, typically a conversation between a user and an agent. While the following description focuses on the user-live agent use case, this is not a limitation. The interactive conversation may also involve some automated process as a participant. Typically, and in the live agent use case, the utterance is a written query or question that the user enters into a conversational chat interface as previously described, although this is not a limitation, as the utterance may be received in the system in other words (orally, email, text message, an input form, or the like)."; Column 19, lines 19-22, "An individual end user typically accesses the system using a user application executing on a computing device (e.g., mobile phone, tablet, laptop or desktop computer, Internet-connected appliance, etc.)."; An interactive conversation between a user and an agent reads on interaction content from an interaction session.),
identify contextual attributes of one or more of the interaction content and the interaction session (Column 17, lines 2-8, "When user input is received, the generative-chat-application first checks the conversation history cache to see if there is any relevant conversation history. If so, the conversation history is passed to the response generation module along with the current user input and the context found from semantic-search to generate a contextually-relevant response."; Finding context from a semantic-search reads on identifying contextual attributes.),
match the interaction content to a subset of vectors within a vector storage [based on labels previously assigned to the subset of vectors] (Column 18, lines 27-29, "As noted, the semantic search receives the utterance from the generative chat application and retrieves the relevant context, in this scenario from the vector database 1514."; Retrieving relevant context from a vector database reads on matching the interaction content to a subset of vectors within a vector storage.),
augment a machine learning (ML) model based on the subset of vectors to generate an augmented ML model (Column 15, lines 52-57, "The semantic-Search API returns the relevant context (when context can be found), and the prompt is then enriched to include it. The resulting enriched prompt is then passed through to a text completion endpoint associated with an external generative AI text completion endpoint, such as OpenAI ChatGPT-3."; Providing an enriched prompt to generative AI reads on augmenting a machine learning model.),
and generate a response for the interaction session based on execution of the augmented ML model on the interaction content and output the response to the source device during the interaction session (Column 15, lines 51-63, "The semantic-Search API returns the relevant context (when context can be found), and the prompt is then enriched to include it. The resulting enriched prompt is then passed through to a text completion endpoint associated with an external generative AI text completion endpoint, such as OpenAI ChatGPT-3. This particular language model is not intended to be limiting, as other large language models may be used for this purpose. Upon receiving a response from the generative AI text completion endpoint, the suggested replies API returns it, e.g., to a front-end application tool that is managing the conversation (namely, the interaction between the user and the live-agent).").
Taylert does not specifically disclose: match the interaction content to a subset of vectors within a vector storage based on labels previously assigned to the subset of vectors.
Penta teaches:
match the interaction content to a subset of vectors within a vector storage based on labels previously assigned to the subset of vectors (Column 3, lines 45-52, "The personalized retrieval-augmented generation system can determine a data context for generating a personalized response by comparing a query embedding with vectorized segments of content items. For instance, the personalized retrieval-augmented generation system determines a data context specific to the entity by comparing the query embedding with a plurality of vectorized segments of content items associated with the entity stored in a database."; Column 18, lines 13-33, "Indeed, the metadata database 406 can store metadata for content items and/or for vectorized segments of content items, where different metadata are associated with or labeled as corresponding to the items or segments. In some instances, metadata can reflect information associated with or generated by an entity. For example, metadata can reflect relationships 410, location(s) 412, and/or the timing 414 associated with content items associated with the entity. For instance, the personalized retrieval-augmented generation system 106 can store metadata about the time of receiving email and who received the email in the metadata database 406. To further illustrate, the personalized retrieval-augmented generation system 106 can receive a query asking for details about an upcoming flight to Nashville. While generated a personalized response about the flight to Nashville, the personalized retrieval-augmented generation system 106 can access metadata regarding the time and date of the upcoming flight to Nashville and utilize the metadata to select relevant information (e.g., vectorized segments of content items) to include in the data context along with a vectorized segment of a flight receipt."; Utilize metadata to select vectorized segments of content items, where the metadata is labeled as corresponding to the items, reads on matching the interaction content to a subset of vectors within a vector storage based on labels previously assigned to the subset of vectors.).
Penta is considered to be analogous to the claimed invention because it is in the same field of machine learning conversational response systems. Therefore, it would have been obvious to someone of ordinary skill in the art before the effective filing date of the claimed invention to have modified Taylert to incorporate the teachings of Penta to utilize metadata to select vectorized segments of content items, where the metadata is labeled as corresponding to the items. Doing so would allow for improving the accuracy of generating responses to queries utilizing retrieval-augmented generation systems or large language models (Penta; Column 4, lines 15-20).
Regarding claim 3, Taylert in view of Penta discloses the apparatus as claimed in claim 1.
Taylert further discloses:
wherein the interaction content comprises chat content from a chat conversation, and the processor is configured to generate a chat response for the chat conversation and output the chat response via a chat window of the chat conversation (Column 8, lines 25-36, "In a typical user case, the software application 402 executes in association with a website 408, although the chatbot functionality may be utilized by multiple distinct websites operated by separate and independent content providers. As such, the computing platform provides the chatbot functionality in a multi-tenant operating environment, although this is not a requirement. The user provides input to the chatbot as speech, as one or more physical actions (e.g., selecting a button or link, entering data in a field, etc.), or as some combination of speech and physical action. In this example, the chatbot 402 is an AI-based conversational bot."; Column 8, lines 51-52, "In the context of a chatbot, the response typically is provided in a chat window.").
Regarding claim 6, Taylert in view of Penta discloses the apparatus as claimed in claim 1.
Taylert further discloses:
wherein the processor is configured to generate a prompt which includes the subset of vectors and a description of a task to be performed by the augmented ML model, and input the prompt to the ML model to generate the augmented ML model (Column 3, lines 45-52, "The personalized retrieval-augmented generation system can determine a data context for generating a personalized response by comparing a query embedding with vectorized segments of content items. For instance, the personalized retrieval-augmented generation system determines a data context specific to the entity by comparing the query embedding with a plurality of vectorized segments of content items associated with the entity stored in a database."; Column 15, lines 52-57, "The semantic-Search API returns the relevant context (when context can be found), and the prompt is then enriched to include it. The resulting enriched prompt is then passed through to a text completion endpoint associated with an external generative AI text completion endpoint, such as OpenAI ChatGPT-3."; Providing an enriched prompt to generative AI reads on augmenting a machine learning model.).
Regarding claim 7, Taylert in view of Penta discloses the apparatus as claimed in claim 1.
Taylert further discloses:
wherein the processor is configured to convert the interaction content into a vector via execution of a second ML model on the interaction content, and match the vector to the subset of vectors (Column 16, lines 25-27, "One model that the Semantic-Search API may use to embed text is a transformer"; Column 18, lines 27-29, "As noted, the semantic search receives the utterance from the generative chat application and retrieves the relevant context, in this scenario from the vector database 1514."; Embedding text using a transformer reads on converting the interaction content into a vector via execution of a machine learning model, and retrieving relevant context from a vector database reads on matching the vector to the subset of vectors.).
Regarding claim 8, arguments analogous to claim 1 are applicable.
Regarding claim 10, arguments analogous to claim 3 are applicable.
Regarding claim 13, arguments analogous to claim 6 are applicable.
Regarding claim 14, arguments analogous to claim 7 are applicable.
Regarding claim 15, arguments analogous to claim 1 are applicable. In addition, Taylert discloses a computer-readable storage medium comprising instructions stored therein (Column 22, lines 26-30, "A machine implementing the techniques herein comprises a hardware processor, and non-transitory computer memory holding computer program instructions that are executed by the processor to perform the above-described methods.") which when executed by a processor cause the processor to perform the steps of claim 1.
Regarding claim 17, arguments analogous to claim 3 are applicable.
Regarding claim 20, arguments analogous to claim 6 are applicable.
Claims 2, 9, and 16 are rejected under 35 U.S.C. 103 as being unpatentable over Taylert in view of Penta, and further in view of Rodriquez et al. (US Patent No. 11,366,857), hereinafter Rodriquez.
Regarding claim 2, Taylert in view of Penta discloses the apparatus as claimed in claim 1, but does not specifically disclose: wherein the processor is configured to identify a policy of an organization which is being discussed during the interaction session and match the interaction content to the subset of vectors based on labels of the subset of vectors including a label of the policy.
Rodriquez teaches:
wherein the processor is configured to identify a policy of an organization which is being discussed during the interaction session and match the interaction content to the subset of vectors based on labels of the subset of vectors including a label of the policy (Column 5, lines 21-28, "The policies storage 103c stores policies (or conditions, criteria, etc.) of an organization that can be used to align transcripts and HCl actions, cluster transcripts, label feature vectors, and train models. For example, it may be a particular company's corporate policy to require certain information to authenticate users prior to some transactions (e.g., a bank may require detailed identification information to authenticate a caller)."; Column 7, lines 3-12, "In several embodiments, the transcripts module 135 tags the transcripts and/or the mapped vectors based on, for example, metadata associated with the transcript (stored in the metadata storage 103j), policies (stored in the policies storage 103c), communication medium/source via which the transcript was received (for example, chat, phone, email, web page, mobile application, smart device, IoT device, social media, etc.), and so on. The transcripts module 135 can store the vector representations of the transcripts and their tags in the transcripts storage 103a."; Tagging a transcript based on policies stored in a policies storage reads on identifying a policy of an organization which is being discussed during the interaction session, and using stored policies to align transcripts and label feature vectors reads on matching the interaction content to the subset of vectors.).
Rodriquez is considered to be analogous to the claimed invention because it is in the same field of machine learning conversational response systems. Therefore, it would have been obvious to someone of ordinary skill in the art before the effective filing date of the claimed invention to have modified Taylert in view of Penta to incorporate the teachings of Rodriquez to tag a transcript based on policies stored in a policies storage and use stored policies to align transcripts and label feature vectors. Doing so would allow for automating tasks that otherwise would require human interaction while simultaneously surfacing best practices among human agents (Rodriquez; Column 3, lines 5-22).
Regarding claim 9, arguments analogous to claim 2 are applicable.
Regarding claim 16, arguments analogous to claim 2 are applicable.
Claims 4 – 5, 11 – 12 and 18 – 19 are rejected under 35 U.S.C. 103 as being unpatentable over Taylert in view of Penta, and further in view of Guo et al. (US Patent No. 9,535,960), hereinafter Guo.
Regarding claim 4, Taylert in view of Penta discloses the apparatus as claimed in claim 1, but does not specifically disclose: wherein the processor is configured to retrieve device data from the source device and identify the plurality of contextual attributes based on execution of one or more additional ML models on the interaction content and the device data.
Guo teaches:
wherein the processor is configured to retrieve device data from the source device and identify the plurality of contextual attributes based on execution of one or more additional ML models on the interaction content and the device data (Column 1, lines 29-34, "A search engine is described herein that retrieves information based, in part, on a context in which a query has been submitted. The search engine operates by using a deep learning model to project context information (associated with the context) into a context concept vector in a semantic space."; Column 1, lines 51-57, "According to one illustrative aspect, the context information may describe text in proximity to the query within a source document, demographic information regarding the user who has submitted the query, the time at which the query was submitted, the location at which the query was submitted, the prior search-related behavior of the user who has submitted the query, etc., or any combination thereof."; Column 5, lines 36-42, "Alternatively, or in addition, the context information may describe the location at which a user has submitted a query. The search engine 112 may determine the location of the user based on any position-determination mechanisms, such as satellite-based mechanisms (e.g., GPS mechanisms), triangulation mechanisms, dead-reckoning mechanisms, and so on. Alternatively, or in addition, the context information may describe the time at which a user has submitted a query."; Retrieving information based on a context in which a query has been submitted, including the location of the user and the time at which the query was submitted, reads on retrieving device data from the source device, and using a deep learning model to project context information into a context concept vector reads on identifying contextual attributes based on execution of a machine learning model.).
Guo is considered to be analogous to the claimed invention because it is in the same field of machine learning conversational response systems. Therefore, it would have been obvious to someone of ordinary skill in the art before the effective filing date of the claimed invention to have modified Taylert in view of Penta to incorporate the teachings of Guo to retrieve information based on a context in which a query has been submitted, including the location of the user and the time at which the query was submitted, and use a deep learning model to project context information into a context concept vector. Doing so would allow for providing more useful search results to a user who has submitted a query (Guo; Column 1, lines 38-50).
Regarding claim 5, Taylert in view of Penta discloses the apparatus as claimed in claim 1, but does not specifically disclose: wherein the processor is configured to identify a geographic location associated with the interaction content and match the interaction content to the subset of vectors based on labels of the subset of vectors including a label of the geographic location.
Guo teaches:
wherein the processor is configured to identify a geographic location associated with the interaction content and match the interaction content to the subset of vectors based on labels of the subset of vectors including a label of the geographic location (Column 1, lines 29-34, "A search engine is described herein that retrieves information based, in part, on a context in which a query has been submitted. The search engine operates by using a deep learning model to project context information (associated with the context) into a context concept vector in a semantic space."; Column 5, lines 36-42, "Alternatively, or in addition, the context information may describe the location at which a user has submitted a query. The search engine 112 may determine the location of the user based on any position-determination mechanisms, such as satellite-based mechanisms (e.g., GPS mechanisms), triangulation mechanisms, dead-reckoning mechanisms, and so on. Alternatively, or in addition, the context information may describe the time at which a user has submitted a query."; Column 6, lines 35-41, "Each transformation module uses an instantiation of the model 106 to map an input vector into an output concept vector. The input vector represents a particular linguistic item, such as a query, context, document, etc. The concept vector is expressed in a semantic space and reveals semantic information regarding the corresponding linguistic item from which it was derived."; Retrieving information based on a context in which a query has been submitted including the location of the user reads on identifying a geographic location associated with the interaction content, and projecting context information into a context concept vector in a semantic space reads on matching the interaction content to the subset of vectors based on labels of the subset of vectors including a label of the geographic location.).
Guo is considered to be analogous to the claimed invention because it is in the same field of machine learning conversational response systems. Therefore, it would have been obvious to someone of ordinary skill in the art before the effective filing date of the claimed invention to have modified Taylert in view of Penta to incorporate the teachings of Guo to retrieve information based on a context in which a query has been submitted including the location of the user, and project context information into a context concept vector in a semantic space. Doing so would allow for providing more useful search results to a user who has submitted a query (Guo; Column 1, lines 38-50).
Regarding claim 11, arguments analogous to claim 4 are applicable.
Regarding claim 12, arguments analogous to claim 5 are applicable.
Regarding claim 18, arguments analogous to claim 4 are applicable.
Regarding claim 19, arguments analogous to claim 5 are applicable.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure:
Gruber et al. (US Patent No. 12,087,308)
Zamani et al. ("Retrieval-Enhanced Machine Learning")
Lewis et al. ("Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks")
Any inquiry concerning this communication or earlier communications from the examiner should be directed to James Boggs whose telephone number is (571)272-2968. The examiner can normally be reached M-F 8:00 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, Daniel Washburn can be reached at (571)272-5551. 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.
/JAMES BOGGS/Examiner, Art Unit 2657