DETAILED ACTION
Introduction
Applicant's submission filed on 01/19/2025 has been entered. Claims 1-20 are pending in the application and have been examined.
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 .
Claim Rejections - 35 USC § 103
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis (i.e., changing from AIA to pre-AIA ) for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
Claims 1-7 and 9-10 are rejected under 35 U.S.C. 103 as being unpatentable over Pathak et. al. US PgPub. 2024/0394477 in view of Griffiths, et. al. US PgPub 2020/0234694.
Regarding claim 1, Pathak teaches a computer implemented method comprising: tagging a user prompted conversation to generate contextually relevant user tags from a user prompt according to pre-populated tags in a Definite Finite Automaton (DFA) tree (see Pathak, [0092, 0094-0097] processes the user query to extract the keyword/NER/topic-modeling ( tagging user prompted conversation) and processing the relevance of the selected terms (according to the pre-populated tags); Pathak [0061] discusses the state data store 144 ( DFA) ); navigating the DFA tree according to tags to determine a user prompted conversation state (see Pathak, [0127] the dialogue system accesses the state data store to provide candidate context information ( determine user prompted conversation state)); accessing dialog identifications of previous conversations that are similar that are associated with the user prompted conversation state (see Pathak, [0127] The candidate context information includes a dialogue history that precedes the input query. The dialogue history, in turn, includes previous input queries submitted to the language model, and previous responses generated by the language model for the previous input queries); compiling an LLM prompt for forming a response to the user prompted conversation(see Pathak, [0127] In block 1510, the dialogue system 104 selects targeted context information (e.g., the targeted context information 202) from the candidate context information by determining a semantic relevance of the input query to each of the plural parts by performing vector-based analysis. In block 1512, the dialogue system 104 creates prompt information (e.g., the prompt information 124) that includes the input query and the targeted context information); and generating a response to the user based on the prompt(see Pathak, [0127] In block 1514, the dialogue system 104 submits the prompt information to the machine-trained language model, and receives a response (e.g., the response 126) from the machine-trained language model based on the prompt information. In block 1516, the dialogue system 104 generates output information (e.g., output information 120) based on the response.).
Pathak teaches tagging a user prompted conversation to generate contextually relevant user tags from a user prompt according to pre-populated tags in a Definite Finite Automaton (DFA) tree based on state of the dialogue as has been processed by the dialogue system, Griffiths teaches tagging a user prompted conversation to generate contextually relevant user tags from a user prompt according to pre-populated tags in a Definite Finite Automaton (DFA) tree (see Griffiths, [0040-0041] describes determining the states from messages in a conversation and creating sequence of states (DFA); Griffiths [0033] describes label ( tags) of the msg) ;navigating the DFA tree according to tags to determine a user prompted conversation state ( see Griffiths, [0084-0085] describes determining the representative conversation for the state and presenting to the user ( determine user prompted conversation state)) ;accessing dialog identifications of previous conversations that are similar that are associated with the user prompted conversation state (see Griffiths, [0080-0081] describes determining similarity between conversations and clustering the conversations accordingly).
Pathak and Griffiths are considered to be analogous to the claimed invention because both relate to dialogue processing. 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 the teachings of Pathak to using machine trained language models to respond to user’s queries with identifying representative or typical conversations from a corpus of conversations by processing the corpus of conversations with a state model teachings of Griffiths to compare conversations in a manner such that conversations about the same subject matter may be recognized as similar even though the conversations have different flows and use different language (see Griffiths, [0005]).
Regarding claim 2, Pathak in view of Griffiths teaches the computer implemented method of claim 1. Pathak further teaches wherein the LLM prompt is formed using in-context learning from the dialog identifications(see Pathak, [0062] discusses the prompt managing component to generate the prompt information based on the query, candidate context information; Pathak [0066] discusses the algorithmic component based on rules( in-context) and the training system based on the training examples ( learning from dialog identification) ; Pathak[0069] discuss the in context learning usage). Griffiths further teaches wherein the LLM prompt is formed using in-context learning from the dialog identifications (see Griffiths, [0041] teaches contextual information added to message embedding ( in-context learning)). The motivation to combine as claim 1 applies here.
Regarding claim 3, Pathak in view of Griffiths teaches the computer implemented method of claim 1. Griffiths further teaches wherein the DFA tree merges tree nodes that have a similarity score Ø sim (q, q'), exceeding a threshold, λ, wherein q is the node is question and q' is the node q is being compared with(see Griffiths, [0065] teaches determining salient scores for each cluster and selecting subset of states for each cluster based on comparing the salient scores to a threshold( merging tree notes); Griffiths, [0080-0081] describes the clustering of conversations). The motivation to combine as claim 1 applies here.
Regarding claim 4, Pathak in view of Griffiths teaches the computer implemented method of claim 1. Griffiths further teaches wherein the conversation's state is determined by δ (Q, Σ), where Q is a finite set of states and Σ is a finite input alphabet(see Griffiths, [0037-0041] describes various word embeddings (finite input alphabet) assigned for the conversation; Griffiths, [0035] describes assigning states to the messages ( finite set of states) ). Pathak also teaches wherein the conversation's state is determined by S (Q, E), where Q is a finite set of states and E is a finite input alphabet(see Pathak, [0059] discusses processing keyword( finite alphabet ) from the source information to create relevant candidate context information (state store, finite state) as described in Pathak [0044]). The same motivation to combine as claim 1 applies here.
Regarding claim 5, Pathak in view of Griffiths teaches the computer implemented method of claim 1. Griffiths further teaches wherein the DFA tree tracks progress of the conversation using an index tracking function, I(qo) wherein I(qo) maps a state q to a set of dialog identifications in { 1 . . . , N}(see Griffiths, [0030-0031, 0066]describes the states of a conversation and the sequence of steps, where the position of the state may be the index of the state in sequence of states( index tracking function)). The same motivation to combine as claim 1 applies here.
Regarding claim 6, Pathak in view of Griffiths teaches the computer implemented method of claim 1. Pathak further teaches wherein the DFA tree merges linguistically different but contextually equivalent tags(see Pathak, [0101] determines group of items that are considered same concept ( contextually equivalent) and then selects one representative of the group ( merge linguistically different but contextually equivalent)). Griffiths further teaches wherein the DFA tree merges linguistically different but contextually equivalent tags ( see Griffiths, [0044, 0053] describes clustering similar messages ( contextually equivalent tags); Griffiths[0064] describes representative sequence of states for the cluster determined by salient selection ( DFA tree merge) ). The same motivation to combine as claim 1 applies here.
Regarding claim 7, Pathak in view of Griffiths teaches the computer implemented method of claim 1. Griffiths further teaches wherein the pre-populated tags in a DFA tree include a complete repository of keywords that describe all possible utterances of an LLM input with a degree of abstraction (see Griffiths, [0075-0076] describes presenting information on each cluster of conversations and a POSITA will determine how to process same for their design of implementation). The same motivation to combine as claim 1 applies here.
Regarding claim 9, is directed to a system claim corresponding to the method claim presented in claim 1 and is rejected under the same grounds stated above regarding claim 1.
Regarding claim 10, is directed to a system claim corresponding to the method claim presented in claim 2 and is rejected under the same grounds stated above regarding claim 2.
Regarding claim 11, is directed to a system claim corresponding to the method claim presented in claim 3 and is rejected under the same grounds stated above regarding claim 3.
Regarding claim 12, is directed to a system claim corresponding to the method claim presented in claim 4 and is rejected under the same grounds stated above regarding claim 4.
Regarding claim 13, is directed to a system claim corresponding to the method claim presented in claim 5 and is rejected under the same grounds stated above regarding claim 5.
Regarding claim 14, is directed to a system claim corresponding to the method claim presented in claim 6 and is rejected under the same grounds stated above regarding claim 6.
Regarding claim 15, is directed to a computer program product claim corresponding to the method claim presented in claim 1 and is rejected under the same grounds stated above regarding claim 1.
Regarding claim 16, is directed to a computer program product claim corresponding to the method claim presented in claim 2 and is rejected under the same grounds stated above regarding claim 2.
Regarding claim 17, is directed to a computer program product claim corresponding to the method claim presented in claim 3 and is rejected under the same grounds stated above regarding claim 3.
Regarding claim 18, is directed to a computer program product claim corresponding to the method claim presented in claim 4 and is rejected under the same grounds stated above regarding claim 4.
Regarding claim 19, is directed to a computer program product claim corresponding to the method claim presented in claim 5 and is rejected under the same grounds stated above regarding claim 5.
Regarding claim 20, is directed to a computer program product claim corresponding to the method claim presented in claim 6 and is rejected under the same grounds stated above regarding claim 6.
Claim 8 is rejected under 35 U.S.C. 103 as being unpatentable over Pathak et. al. US PgPub. 2024/0394477 in view of Griffiths, et. al. US PgPub 2020/0234694 further in view of Ozonat et.al US PgPub. 2014/0040297.
Regarding claim 8, Pathak in view of Griffiths teaches the computer implemented method of claim 1. However, Pathak in view of Griffiths fail to teach wherein the tagging generates no more than three words per tag.
However, Ozonat teaches wherein the tagging generates no more than three words per tag (see Ozonat, [0031] describes determining tags based on words/phrases ( two or mode) relevant to the domain of interest; the determination of no more than three words per tag is a design choice of implementation ).
Pathak in view of Griffiths teach processing prior user conversations to determine the response to user queries, however does not teach length of tags or labels. Ozonat teaches determining tags based on two or more words relevant to the domain of interest. Using the known technique of determining tags of particular length based on the content as taught by Ozonat (see Ozonat, [0031]), to provide the length of tags in the references Pathak in view of Griffiths, such that design choice of the particular length of tag as claimed would have been obvious to one of ordinary skill in the art.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
Sun, Y, et al. "DFA-RAG: Conversational semantic router for large language model with definite finite automaton." arXiv preprint arXiv:2402.04411 (2024) teaches the retrieval-augmented large language model with Definite Finite Automaton (DFA-RAG), a novel framework designed to enhance the capabilities of conversational agents using large language models (LLMs) (see Sun, section 1).
Yuan et al US Patent 12,639,529 teaches a semantic retrieval component that can extract information from an online source, according to a query, to generate an in-context learning input utilized by an LLM for responding to the query (see Yuan, abstract).
Any inquiry concerning this communication or earlier communications from the examiner should be directed to NANDINI SUBRAMANI whose telephone number is (571)272-3916. The examiner can normally be reached Monday - Friday 12:00pm - 5:00 pm EST.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Bhavesh M Mehta can be reached at (571)272-7453. 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.
/NANDINI SUBRAMANI/ Examiner, Art Unit 2656
Prosecution Insights