DETAILED ACTION
This communication is in response to the Amendments and Arguments filed on 26 January 2026. Claims 1-20 are pending and have been examined. The Applicants’ amendment and remarks have been carefully considered, but are not persuasive. Hence, this Action has been made FINAL.
All previous objections and rejections directed to the Applicant’s disclosure and claims not discussed in this Office Action have been withdrawn by the Examiner.
Response to Amendments and Arguments
Applicant's arguments with respect to the 101 and art rejections have been fully considered, but they are not persuasive.
The applicant asserts that the claims are not abstract at least because they recite elements that are incapable of performance in the mind. However, as explained in the non-final rejection, the additional elements are recited only at a high level of generality. Also, terms such as “generative model”, “intent classification module”, and “anchor generation module” are not recited with any details as to how they are implemented. Thus, at the highest level of generality, the steps of the claims may be performed mentally. The applicant further asserts that the claims recite elements that integrate any alleged abstract idea recited in such claims into a practical application of the abstract idea, specifically that the claims aim to improve the functioning of a computer or an improvement to other suitable technology or technical field, such as generative models, by improving the content that is presented to a user in connection with conversational data resultant from interaction with the generative model. However, the examiner observes that the steps “receiving”, “obtaining”, and “transmitting” steps are mere data gathering and output recited at a high level of generality, and thus are insignificant extra-solution activity. The “providing” steps are recited at a high level of generality and amount to no more than mere instructions to apply the exception using a generic computer. No details are provided about the how the outcomes are accomplished. In other words, and as stated earlier, no details are provided as to how the “generative model”, “intent classification module”, and “anchor generation module” are implemented and achieve their respective outcomes. Hence, the 101 rejections are maintained.
The applicant states that Lakshmikanthan does not teach providing both "input set forth by a user" and "output generated by a generative model responsive to receiving the input." Nothing in Lakshmikanthan discloses conversational data that comprises an output generated by a generative model responsive to receiving the input. Lakshmikanthan, para [0018] and fig. 2, explain that outputs of the models can include can include predicted intent classifications and confidence values. And, para [0052], explains that for each user query, the intent-based user query router 511 obtains a predicted intent class or intent classification from a LLM fine-tuned for intent classification for a user query. The LLM is a generative model. Additionally, Lakshmikanthan, para [0014], teaches that the language model-based intent classifier 115 can be a transformer-based LLM. Thus, the output (i.e., predicted intent classifications (217)) is generated by a generative model responsive to receiving the input (i.e., user query).
The applicant further asserts that Lakshmikanthan only aims to determine a user intent - not anchor text. Anchor text, as discussed in the specification, is indicative of portions of the conversational data that are correlated with the user intent, which is clearly distinguishable from an intent alone. The examiner refutes this assertion. As noted in the non-final rejection, the examiner has interpreted the query-based prompt as anchor text. With reference to para [0022] and fig. 2 of Lakshmikanthan, at stage B the intent-based query router 101 generates prompt 211 based on the user query 205 for intent classification…As an example, the intent-based query router 101 may analyze the conversation history to determine topic transition. Conversation history data C1 and intent class 217 are fed to the intent-based query router 101 (i.e., an anchor generation module) to generate the query-based prompt (i.e., anchor text). Thus the anchor text (i.e., query-based prompt) is based on input conversation data that are correlated/associated with the user intent. Hence, the 102 and 103 rejections are maintained.
Claim Rejections - 35 USC § 101
35 U.S.C. 101 reads as follows:
Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.
Claims 1-20 are rejected under 35 USC 101 because the claimed invention is directed to an abstract idea without significantly more. The claims recite steps for optimizing content delivery using intent classification and query refinement. The limitations of claims 1-20, as drafted, are a computer program product or system that, under their broadest reasonable interpretation, cover performance of the limitations in the mind but for the recitation of generic computer components. That is, other than reciting “computer readable storage medium”, “program instructions” “computer”, “processor”, and “memory” nothing in the claim element precludes the steps from practically being performed in the mind and/or with pen and paper calculations. Under the BRI, a human could manually feed data into a ML model. Furthermore, aside from being generic, the ML model does not actively perform any of these steps. The feeding step and the extracting step could just involve a human mentally considering a question/prompt via reading. There appear to be no technical specifics about how this ML model might be structured or carry out analysis of the prompt and query data. Accordingly, the steps of the claims are directed to organizing human interactions and/or a mental process. If a claim limitation, under its broadest reasonable interpretation, covers performance of the limitation in the mind but for the recitation of generic computer components, then it falls within the “Mental Processes” grouping of abstract ideas. Accordingly, the claims recite an abstract idea.
This judicial exception is not integrated into a practical application. In particular, claims 1-20 only recite the additional elements “computer readable storage medium”, “program instructions” “computer”, “processor”, and “memory” to perform the aforementioned steps. The processor and other hardware are recited at a high-level of generality (i.e., as a generic processor performing a generic computer function for transliterating text such that they amount to no more than mere instructions to apply the exception using generic computer components.
The claims do not include additional elements that are sufficient to amount to significantly more than the judicial exception. As discussed above with respect to integration of the abstract idea into a practical application, the additional hardware elements to perform both the aforementioned steps amount to no more than mere instructions to apply the exception using generic computer components. Mere instructions to apply an exception using generic computer components cannot provide an inventive concept. The claims are not patent eligible.
Claim Rejections - 35 USC § 102
The following is a quotation of the appropriate paragraphs of 35 U.S.C. 102 that form the basis for the rejections under this section made in this Office action:
A person shall be entitled to a patent unless –
(a)(2) the claimed invention was described in a patent issued under section 151, or in an application for patent published or deemed published under section 122(b), in which the patent or application, as the case may be, names another inventor and was effectively filed before the effective filing date of the claimed invention.
Claim(s) 1-3, 6-7, 10-12, 14-16, 18, and 19-20 is/are rejected under 35 U.S.C. 102(a)(2) as being anticipated by US 20250068857, hereinafter referred to as Lakshmikanthan et al.
Regarding claim 1 (Original), Lakshmikanthan et al. discloses a computing system comprising:
a processor (Lakshmikanthan et al., fig. 5(501).); and
memory (Lakshmikanthan et al., fig. 5(507).) storing instructions that, when executed by the processor, cause the processor to perform acts (Lakshmikanthan et al., para [0048]-[0051].) comprising:
receiving conversational data comprising an input set forth by a user of a client computing device and output generated by a generative model responsive to receiving the input (“The intent-based query router 101 submits a prompt to the language model-based intent classifier 115 via the intent classification interface 105 and obtains predicted intent classifications from the intent classifier 115 via the intent classification interface 105…The language model-based intent classifier 115 can be a transformer-based LLM but can be other text generation tools tuned for intent classification,” Lakshmikanthan et al., para [0014]. Here, the conversational data includes a query-based prompt (i.e., an input to the language model-based intent classifier) and an intent classification (i.e., an output to the language model-based intent classifier). The LLM is a generative model.);
providing the conversational data to an intent classification module (Lakshmikanthan et al., fig. 1(115).), wherein the intent classification module produces an output indicative of a user intent based upon the conversational data (“In FIG. 2, the language model-based intent classifier 115 is depicted with multiple models. Outputs of the models can include predicted intent classifications and confidence values,” Lakshmikanthan et al., para [0018].);
providing the conversational data and the output indicative of a user intent as input into an anchor generation module (Lakshmikanthan et al., Stage C1, serves as an anchor generation module.), wherein the anchor generation module generates anchor text based upon the conversational data and the output (“At stage C1, the intent-based query router 101 accesses the conversation history repository 107 to analyze conversation history. As an example, the intent-based query router 101 may analyze the conversation history to determine topic transition. In addition, the intent-based query router 101 updates state of the conversation to reflect receipt of an utterance or query from the user. Based on conversation history, the intent-based query router 101 may add an indication of a current topic to the query 205 to form the prompt 211 updated query?,” Lakshmikanthan et al., para [0022]. As noted, the intent classification is the output (of the generative model), and the conversation history is conversational data, both of which are used to generate the query-based prompt (i.e., anchor text).), wherein the anchor text is indicative of portions of the conversational data correlated with the user intent (Lakshmikanthan et al., para [0022]. The query-based prompt (i.e., anchor text) is derived from conversation history (i.e., conversational data) correlated with intent classification (i.e., user intent).);
generating a content query based upon the anchor text (Lakshmikanthan et al., fig. 4(403).);
obtaining content responsive to the content query (Lakshmikanthan et al., fig. 4(405).); and
transmitting the content to the client computing device for presentation to the user (“FIG. 4 is a flowchart of example operations for routing query responses back to users,” Lakshmikanthan et al., para [0007].).
As to claim 14, claim 14 is rejected on the same grounds as claim 1.
As to claim 18, CRM claim 18 and system claim 1 are related as system and CRM of using same, with each claimed element’s function corresponding to the system step. Accordingly claim 18 is similarly rejected under the same rationale as applied above with respect to system claim. Also, Lakshmikanthan et al., para [0052] teaches a processor and associated computer-implemented program product.
Regarding claim 2 (Original), Lakshmikanthan et al. discloses the computing system of claim 1, wherein generating the content query based upon the anchor text comprises:
generating a content query generation prompt (As seen in Lakshmikanthan et al., fig. 2(112)(213)(D), intent classification and conversation history are used to generate the content query generation prompt.), wherein the content query generation prompt comprises the anchor text (Lakshmikanthan et al., para [0022]. As noted, the intent classification is the output (of the generative model), and the conversation history is conversational data, both of which are used to generate the query-based prompt (i.e., anchor text).) and user profile information associated with the user (“ For example, the intent-based query router 101 may obtain role of a digital identity associated with a conversation and current products and/or services of a customer and store the information into the conversation context repository 109,” Lakshmikanthan et al., para [0013]. Digital identity is a type of user profile information.); and
providing the content query generation prompt as input into an instance of the generative model (“The intent-based query router 101 submits a prompt to the language model-based intent classifier 115 via the intent classification interface 105,” Lakshmikanthan et al., para [0014]. The intent classification interface 105 serves as a generative model interface since the language model-based intent classifier 115 is a generative model (i.e., LLM). See also Lakshmikanthan et al., fig. 2(211)(115).); and
receiving, from the instance of the generative model, an output comprising the content query (As noted with respect to claim 1, the generative model outputs an intent classification, which includes the query.).
As to claim 15, claim 15 is rejected on the same grounds as claim 2.
As to claim 19, CRM claim 19 and system claim 2 are related as system and CRM of using same, with each claimed element’s function corresponding to the system step. Accordingly claim 19 is similarly rejected under the same rationale as applied above with respect to system claim. Also, Lakshmikanthan et al., para [0052] teaches a processor and associated computer-implemented program product.
Regarding claim 3 (Original), Lakshmikanthan et al. discloses the computing system of claim 1, the acts further comprising:
at the intent classification module, generating an intent classification prompt, wherein the intent classification prompt comprises the conversational data and user profile information associated with the user (Lakshmikanthan et al., fig. 2(213)(D), serves as an intent classification prompt which comprises conversational data (from C1) and user profile information (from C2).);
providing the intent classification prompt as input into an instance of the generative model associated with the intent classification module (Lakshmikanthan et al., fig. 2(211)(115), shows that the prompt is input to the LLM 115 (i.e., generative model).); and
receiving, from the instance of the generative model, an output indicative of a user intent, wherein the output is based upon the conversational data and the user profile information (The LLM 115 outputs an intent classification (i.e., user intent).).
As to claim 16, claim 16 is rejected on the same grounds as claim 3.
Regarding claim 6 (Original), Lakshmikanthan et al. discloses the computing system of claim 1, wherein obtaining the content responsive to the content query comprises:
generating a content generation prompt based upon the content query and a content asset obtained from a content asset data store (“The process of obtaining an intent classification based on a user query can vary depending upon implementation, the user query, and state of the conversation. At stage B, the intent-based query router 101 generates prompt 211 based on the user query 205 for intent classification. The prompt 211 may be the query 205 or may be a modified version of the query 205. For instance, the intent-based query router 101 may pre-process the query 205 to lint the query 205 which yields the prompt 211…Based on conversation history, the intent-based query router 101 may add an indication of a current topic to the query 205 to form the prompt 211. At stage C2, the intent-based query router 101 determines whether additional text should be added to the query 205 to form the prompt 211 to provide context to the intent classifier 115. The context information may be used to disambiguate the query 205 and form the prompt 211,” Lakshmikanthan et al., para [0022]. See also Lakshmikanthan et al., fig. 2(211). Is interpreted as a content asset.); and
providing the content generation prompt as input into an instance of the generative model (“At stage D, the intent-based query router 101 submits the formed prompt 211 to the queue 213 from which the intent classifier 115 retrieves prompts for predicting intent classifications,” Lakshmikanthan et al., para [0023]. As already noted, the intent classifier 115 operates via an LLM (i.e., generative model). See also Lakshmikanthan et al., fig. 2(115)(213).); and
receiving, from the instance of the generative model, an output comprising content responsive to the content query (“In FIG. 2, the language model-based intent classifier 115 is depicted with multiple models. Outputs of the models can include predicted intent classifications and confidence values,” Lakshmikanthan et al., para [0018]. The predicted intent classifications (outputs) comprise content responsive to the content query – i.e., recognizing the intent classification to which the content query belongs.).
As to claim 20, CRM claim 20 and system claim 6 are related as system and CRM of using same, with each claimed element’s function corresponding to the system step. Accordingly claim 20 is similarly rejected under the same rationale as applied above with respect to system claim. Also, Lakshmikanthan et al., para [0052] teaches a processor and associated computer-implemented program product.
Regarding claim 7 (Original), Lakshmikanthan et al. discloses the computing system of claim 1, the acts further comprising:
wherein when the content obtained responsive to the content query comprises a plurality of content, determining a score representative of a likelihood that a content will be interacted with at the client computing device (“The intent-based query router may rank/prioritize the query responses and select a top n of the query responses according to the ranking. As an example, the backend services may be associated with values used for prioritizing and ranking. For example, a live agent may have a higher priority or ranking value than a response from a conversational agent training on troubleshooting manual. Customer attributes may also be a factor in the ranking. For instance, a customer may have multiple licenses with different expirations dates. The expiration date and quantity of each license can inform the ranking of query responses,” Lakshmikanthan et al., para [0043]. The ranking is interpreted as a score. Top ranked query responses incorporating customer attributes are more likely to be interacted with at the client device.);
selecting the content with the highest score (Lakshmikanthan et al., para [0043]. It is an evident design feature that the content with the highest score is selected.); and
transmitting the content with the highest score to the client computing device (Lakshmikanthan et al., para [0043]. It is an evident design feature that the content with the highest score is selected.).
Regarding claim 10 (Original), Lakshmikanthan et al. discloses the computing system of claim 1, wherein the conversational data is received from the generative model (“A chatbot is one type of conversational agent. While the underlying algorithms for chatbots can vary, many chatbots now use large language models (LLMs), especially generative pre-trained LLMs,” Lakshmikanthan et al., para [0010]. And, “In FIG. 2, the language model-based intent classifier 115 is depicted with multiple models. Outputs of the models can include predicted intent classifications and confidence values,” Lakshmikanthan et al., para [0018]. The predicted intent classifications (outputs) is conversational data – i.e., classification derived from the conversation(s).).
Regarding claim 11 (Original), Lakshmikanthan et al. discloses the computing system of claim 10, wherein the conversational data comprises data from a first interaction between the client computing device and the generative model and a second interaction between the client computing device and the generative model, wherein the first interaction preceded the second interaction (“The intent-based query router 101 uses the repositories 107, 109 to store information for tracking conversation state, possibly disambiguate queries, and edit responses across multiple sessions of conversations with different users. The intent-based query router 101 uses the conversation history repository 107 and the conversation context repository 109 to track state transitions in a conversation and context data for a conversation, respectively. Conversation history can be limited to capability of the repository 107 (e.g., a rolling log of queries and responses up to a threshold number of tokens that can be accommodated by the repository 107),” Lakshmikanthan et al., para [0013]. The conversation history is rolling log a queries and responses comprising conversational data from a first interaction between the client computing device and the generative model and a second interaction between the client computing device and the generative model, wherein the first interaction preceded the second interaction.).
Regarding claim 12 (Original), Lakshmikanthan et al. discloses the computing system of claim 10, wherein the content is presented within a generative model interface (“The intent-based query router 101 submits a prompt to the language model-based intent classifier 115 via the intent classification interface 105,” Lakshmikanthan et al., para [0014]. The intent classification interface 105 serves as a generative model interface since the language model-based intent classifier 115 is a generative model (i.e., LLM).).
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) 4 and 17 is/are rejected under 35 U.S.C. 103 as being unpatentable over US 20250068857, hereinafter referred to as Lakshmikanthan et al., in view of US 20210158815, hereinafter referred to as Lee.
Regarding claim 4 (Original), Lakshmikanthan et al. discloses the computing system of claim 1, wherein the intent classification module comprises a sequence-to-sequence model configured to produce the output indicative of a user intent based upon the conversational data (“The intent discriminator 262 may apply an intent classification learning model based on a natural language processing artificial neural network to the first input…In the present embodiment, the intent classification learning model based on the natural language processing artificial neural network may be a sequence-sequence model using a bidirectional LSTM network, and may fill a slot and may simultaneously predict the intent. For example, when the sentence “find orange T-shirt boy” is given, a task may appropriately output or fill slots {action: find people} and {who: boy},” Lee, para [0139].). Lee benefits Lakshmikanthan et al. applying a bidirectional LSTM to enhance contextual understanding of the text. Therefore, it would be obvious for one skilled in the art to combine the teachings of Lakshmikanthan et al. with those of Lee to extract contextual information, thereby improving the intent classification of Lakshmikanthan et al.
As to claim 17, claim 17 is rejected on the same grounds as claim 4.
Claim(s) 5 is/are rejected under 35 U.S.C. 103 as being unpatentable over US 20250068857, hereinafter referred to as Lakshmikanthan et al., in view of, in view of US 20170061515, hereinafter referred to as Hummel.
Regarding claim 5 (Original), Lakshmikanthan et al. discloses the computing system of claim 1, but not wherein obtaining content responsive to the content query comprises: receiving content provided from a content server, wherein the content provided from the content server is obtained based upon a content auction at the content server. Hummel is cited to disclose wherein obtaining content responsive to the content query comprises: receiving content provided from a content server, wherein the content provided from the content server is obtained based upon a content auction at the content server (“When selecting the content item for display, the data processing system (e.g., content server, contextual serving system, advertisement server) receives bids from numerous content providers (e.g., advertisers) and enters them into an online auction (e.g., an online advertisement bidding auction),” Hummel, para [0013].). Hummel benefits Lakshmikanthan et al. by determining an allocation or price of an electronic content item (e.g., electronic online advertisement) to be displayed alongside organic, non-advertisement search result links (Hummel, para [0014]). Therefore, it would be obvious for one skilled in the art to combine the teachings of Lakshmikanthan et al. with those of Hummel to optimize content display of Lakshmikanthan et al.
Claim(s) 8-9 is/are rejected under 35 U.S.C. 103 as being unpatentable over US 20250068857, hereinafter referred to as Lakshmikanthan et al., in view of US 20250077511, hereinafter referred to as Zhao et al.
Regarding claim 8 (Original), Lakshmikanthan et al. discloses the computing system of claim 1, but not the acts further comprising: causing anchor text within the conversational data to be graphically emphasized when presented at the client computing device. Zhao et al. is cited to disclose the acts further comprising: causing anchor text within the conversational data to be graphically emphasized when presented at the client computing device (“The chatbot system presents the results and the summary to the user as a response to the provided query,” Zhao et al., para [0012]. The presentation of a summary is interpreted as an emphasis of the conversational exchange.). Zhao et al. benefits Lakshmikanthan et al. by providing the user with a visual information summary, thereby allowing the user quick ingestion of relevant conversational data. Therefore, it would be obvious for one skilled in the art to combine the teachings of Lakshmikanthan et al. with those of Zhao et al. to improve the conversational data display of Lakshmikanthan et al.
Regarding claim 9 (Original), Lakshmikanthan et al., as modified by Zhao et al., discloses the computing system of claim 8, the acts further comprising:
causing the content to be presented to the user at the client computing device upon a user interaction with the anchor text, wherein the content is presented as a graphical overlay at the client computing device (Zhao et al., para [0012].).
Claim(s) 13 is/are rejected under 35 U.S.C. 103 as being unpatentable over US 20250068857, hereinafter referred to as Lakshmikanthan et al., in view of US 20250036695, hereinafter referred to as De Barros et al.
Regarding claim 13 (Original), Lakshmikanthan et al. discloses the computing system of claim 10, but not wherein the content is supplemental to webpage content concurrently displayed at the client computing device De Barros et al. is cited to disclose wherein the content is supplemental to webpage content concurrently displayed at the client computing device (“In another example, the additional information can be obtained from a web browser (or other application) that has loaded a webpage being viewed by a user,” De Barros et al., para [0025].). De Barros et al. benefits Lakshmikanthan et al. by allowing additional conversational information to be viewed and analyzed by a user, thereby helping the user to make better decisions. Therefore, it would be obvious for one skilled in the art to combine the teachings of Lakshmikanthan et al. with those of De Barros et al. to improve the conversational data display of Lakshmikanthan et al.
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 ANNE L THOMAS-HOMESCU whose telephone number is (571)272-0899. The examiner can normally be reached on Mon-Fri 8-6.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Bhavesh Mehta can be reached on 5712727453. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of an application may be obtained from the Patent Application Information Retrieval (PAIR) system. Status information for published applications may be obtained from either Private PAIR or Public PAIR. Status information for unpublished applications is available through Private PAIR only. For more information about the PAIR system, see http://pair-direct.uspto.gov. Should you have questions on access to the Private PAIR system, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative or access to the automated information system, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/ANNE L THOMAS-HOMESCU/Primary Examiner, Art Unit 2656
Prosecution Insights