DETAILED ACTION
Status of the Claims
This office action is submitted in response to the amendment filed on 6/3/26.
Examiner notes Applicant’s foreign priority date of 11/14/24, which stems from JP2024-198859.
Claims 1-3 and 5-11 have been amended.
Claims 12-14 are new.
Therefore, claims 1-14 are currently pending 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 § 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 1-14 are rejected under 35 U.S.C. § 101 because the claimed invention is directed to a judicial exception (i.e., a law of nature, a natural phenomenon, or an abstract idea) without significantly more.
Step 1: Claims 1-14 are directed to patent-eligible subject matter categories under 35 U.S.C. § 101. See MPEP § 2106.03. Specifically, claim 1 recites a terminal apparatus comprising a processor, and thus falls within the “machine” category. Claim 10 recites a control method that is implemented by a terminal apparatus, and thus falls within the “process” category. Claim 11 recites a non-transitory computer readable storage medium, and thus falls within the “article of manufacture” category. Claims 2-9 and 12-14 depend from claim 1, and thus likewise fall within the “machine” category. Accordingly, the claims satisfy Step 1.
Step 2A, Prong One: Independent claims 1, 10, and 11, in part, describe an invention comprising: (1) extracting a search query from conversation information on a user in a chat service; (2) generating an advertisement link for displaying an advertisement based on the extracted search query; and (3) extracting, from the conversation information, a search query by which an advertisement that is highly likely to achieve conversion is displayed, based on a search query that was input by the user in the past when conversion was achieved with respect to an advertisement displayed in search advertising. As such, the invention is directed to the abstract idea of analyzing user conversation data and targeting ads to the user (based on said conversation data) that are most likely to result in conversion, which, pursuant to MPEP § 2106.04(a), is aptly categorized as a method of organizing human activity (i.e., targeted commercial advertising and marketing) and a mental process (the core steps of identifying a topic of interest from a conversation and selecting an advertisement likely to result in conversion can be performed in the human mind or with pen and paper: an advertising professional can read a chat conversation, recall which of the customer's past inquiries led to purchases, and select an appropriate advertisement and corresponding link for the user, without the aid of a computer. See CyberSource Corp. v. Retail Decisions, Inc., 654 F.3d 1366, 1373 (Fed. Cir. 2011); Elec. Power Grp., LLC v. Alstom S.A., 830 F.3d 1350, 1354 (Fed. Cir. 2016)). Therefore, under Step 2A, Prong One, the claims recite a judicial exception.
Next, the aforementioned claims recite additional elements that are associated with the judicial exception, including: displaying the generated advertisement link in the chat service (claims 1, 10, and 11); making a request to the information processing apparatus for an advertisement content and displaying the advertisement content that is transmitted from the information processing apparatus in accordance with the request (claim 2); displaying a search result content (claims 3 and 6); displaying the advertisement link in a talk room (claims 12 and 13); and storing a search result that was displayed in the past and making a request to an information processing apparatus (claim 14). Examiner understands these limitations to be insignificant extra-solution activity, i.e., data storage, data transmission, and output display at the periphery of the claims, that does not transform the nature of the claims. See Accenture Global Servs., GmbH v. Guidewire Software, Inc., 728 F.3d 1336, 108 U.S.P.Q.2d 1173 (Fed. Cir. 2013), citing Cf. Diamond v. Diehr, 450 U.S. 175, 191-192 (1981) (“[I]nsignificant post-solution activity will not transform an unpatentable principle into a patentable process.”).
The aforementioned claims also recite additional elements including “a terminal apparatus,” “a processor,” “an information processing apparatus,” and “a non-transitory computer readable storage medium.” These limitations are recited at a high level of generality and appear to be nothing more than generic computer components used to apply the abstract idea. Claims that amount to nothing more than an instruction to apply the abstract idea using a generic computer do not render an abstract idea eligible. Alice Corp. Pty. Ltd. v. CLS Bank Int'l, 573 U.S. 208, 223 (2014).
The aforementioned claims further recite that the processor stores therein a “Large Language Model (LLM)” that is used to perform the extraction (claims 1, 10, and 11), and dependent claim 9 further recites the use of a “model” that is trained to output the search query while adopting the conversation information as input. The use of a machine learning model, including an LLM, to receive conversation information as input, process that data, and output a search query amounts to mere instructions to implement the abstract idea on a computer, merely using the model as a tool to perform the abstract data analysis. See MPEP § 2106.05(f); Alice Corp., 573 U.S. at 223; Recentive Analytics, Inc. v. Fox Corp., No. 2023-2437 (Fed. Cir. Apr. 18, 2025) (application of generic machine learning to a new data environment, without improvement to the machine learning models themselves, does not confer eligibility). That the LLM is trained on a particular selection of data (i.e., the user's past conversion-achieving search queries) does not alter this analysis; the claims do not purport to improve how the LLM itself operates, but merely apply a generically-recited LLM, trained on the selected data, as a tool to carry out the extraction.
Step 2A, Prong Two: Furthermore, looking at the elements individually and in combination, the claims as a whole do not integrate the judicial exception into a practical application because they fail to: improve the functioning of a computer or a technical field, apply the judicial exception in the treatment or prophylaxis of a disease, apply the judicial exception with a particular machine, effect a transformation or reduction of a particular article to a different state or thing, or apply the judicial exception beyond generally linking the use of the judicial exception to a particular technological environment. Rather, the claims merely use a computer as a tool to perform the abstract idea(s), and/or add insignificant extra-solution activity to the judicial exception, and/or generally link the use of the judicial exception to a particular technological environment (e.g., a generic computer running a chat service). Accordingly, the claims do not integrate the judicial exception into a practical application, and the analysis proceeds to Step 2B.
Step 2B: Next, the claims do not include additional elements sufficient to amount to significantly more than the judicial exception because the additional elements, when considered both individually and as an ordered combination, do not amount to significantly more than the abstract idea. Furthermore, looking at the limitations as an ordered combination adds nothing that is not already present when looking at the elements taken individually. Simply put, there is no indication that the combination of elements improves the functioning of a computer (or any other technology), and their collective functions are merely facilitated by generic computer implementation.
Additionally, pursuant to the requirement under Berkheimer v. HP Inc., 881 F.3d 1360, 125 USPQ2d 1649 (Fed. Cir. 2018), the following citations are provided to demonstrate that the extra-solution additional elements amount to activities that are well-understood, routine, and conventional. See MPEP § 2106.05(d).
Displaying an advertisement or content to a user via a user interface. Ultramercial, Inc. v. Hulu, LLC, 772 F.3d 709, 715-16, 112 USPQ2d 1750, 1755 (Fed. Cir. 2014).
Receiving or transmitting data over a network. Intellectual Ventures I LLC v. Symantec Corp., 838 F.3d 1307, 1321, 120 USPQ2d 1353 (Fed. Cir. 2016) (utilizing an intermediary computer to forward information); OIP Techs., Inc. v. Amazon.com, Inc., 788 F.3d 1359, 1363, 115 USPQ2d 1090, 1093 (Fed. Cir. 2015) (sending messages over a network).
Storing and retrieving information in memory. Versata Dev. Group, Inc. v. SAP Am., Inc., 793 F.3d 1306, 1334, 115 USPQ2d 1681, 1701 (Fed. Cir. 2015); OIP Techs., 788 F.3d at 1363.
Thus, taken alone and in combination, the additional elements do not amount to significantly more than the above-identified judicial exception (the abstract idea), and are ineligible under 35 U.S.C. § 101.
Dependent claims 2-9 and 12-14 depend from independent claim 1 and include all the limitations contained therein. These claims do not recite any additional technical elements beyond those addressed above, and simply disclose additional limitations that further limit the abstract idea. Specifically:
Claims 2-3 further specify requesting an advertisement content upon the user's selection of the advertisement link and displaying a search result content that includes the advertisement content and a search result based on the search query. These limitations further define the content gathered and output in furtherance of the targeted advertising, and the requesting, receiving, and displaying acts constitute the generic extra-solution activity addressed above.
Claims 4-6 further specify the timing and source of the search results, including a search result retrieved based on the search query in the past by the user (claim 4), a search result retrieved when the information processing apparatus receives a request for the advertisement content (claim 5), and display at a timing that is designated by the user (claim 6). These timing parameters and data-source details further define the abstract idea.
Claims 7-8 further specify that the processor shares the search result content with a conversation partner of the user in the chat service, including when the user gives a permission (claim 8). Sharing content between the user and the conversation partner, conditioned on the user's permission, constitutes managing interactions between people and further defines the certain method of organizing human activity.
Claim 9 further recites the use of a model that is trained to output the search query while adopting the conversation information as input. As set forth above, this limitation amounts to a mere instruction to apply the abstract idea using a machine learning model as a tool. See MPEP § 2106.05(f).
Claim 12 further specifies that the advertisement link is displayed as utterance of a third person other than the user and the conversation partner and for only the user between the user and the conversation partner. This advertisement presentation strategy, directed to how and to whom the advertisement is presented, further defines the abstract idea of targeted advertising, and the display act itself constitutes the generic extra-solution activity addressed above.
Claim 13 further specifies that the advertisement link is displayed at a timing at which a conversation between the user and a conversation partner is paused, wherein the timing is a timing after a lapse of a predetermined time or more since last utterance. These timing parameters further define the abstract idea.
Claim 14 further specifies that the processor stores therein a search result that was displayed in response to input of the search query by the user in the past, and makes a request to an information processing apparatus for only an advertisement that is displayed based on the search query. The determination to request only the advertisement based on the stored past result further defines the abstract idea, and the storing and requesting acts constitute the generic extra-solution activity addressed above.
Therefore, claims 1-14 are not drawn to eligible subject matter, as they are directed to an abstract idea without significantly more.
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 and 9-11 are rejected under 35 U.S.C. § 103 as being unpatentable over Ellis et al. (US 2022/0366454 A1) in view of Rofouei et al. (US 2024/0289407 A1) in view of Alakuijala et al. (US 2023/0281193 A1) and further in view of Abdallah et al. (US 11,144,811 B2).
Claims 1 and 10-11: Ellis discloses a terminal apparatus, control method, and non-transitory computer readable storage medium comprising:
extracting a search query from conversation information on a user in a chat service (Paragraph 0042: “The chat service may be further configured to programmatically determine user intent from one or more chat users of a mobile chat session to extract search terms for targeted promotion selection”; Paragraph 0081: “chat service circuitry 210 may receive input chat data from the mobile chat sessions generated by connected consumers, and may extract search terms and/or intent from the input chat data”; Fig. 8, elements 806, 810, 814, illustrating extraction of search terms “Restaurant,” “Palo Alto,” and “Italian” directly from user messages in the chat session.);
generating an advertisement link for displaying an advertisement based on the extracted search query (Paragraph 0042: the universal relevance service is “configured to receive search terms and provide targeted promotion scoring and/or ranking”; Paragraph 0081: “the chat service circuitry 210 may select targeted promotions and provide electronic marketing communications in the form of output chat data to the mobile chat session”; Fig. 8, element 816, showing “Promotion 1 [Link], Promotion 2 [Link], Promotion 3 [Link]” generated in direct response to the search terms extracted from the chat session.); and
displaying the generated advertisement link in the chat service (Paragraph 0042: the chat service is “configured to interface with mobile chat applications to provide electronic marketing communications as chat messages”; Paragraph 0063: “The server 104 may facilitate the generation and providing of various electronic marketing communications, including output chat data for mobile chat sessions”; Fig. 8, element 816, showing the generated promotion links displayed within the message display 802 of the live mobile chat session.)
Ellis does not appear to explicitly disclose wherein the processor stores therein a Large Language Model (LLM) that performs the extraction, or wherein the processor extracts, from the conversation information by using the trained LLM, a search query (claims 10 and 11 recite commensurate limitations wherein the extracting uses the LLM).
Rofouei, however, discloses wherein the processor stores therein a Large Language Model (LLM) and wherein the processor extracts, from the conversation information by using the... LLM, a search query (Paragraph 0007: a first large language model (LLM) processes data indicative of the user's state, such as the query and/or contextual information, to generate LLM output; Paragraph 0020: the contextual information includes one or more prior queries issued by the user during the search session and data extracted from responsive content returned during the session; Paragraph 0019: one or more synthetic queries are generated using the LLM output; Paragraph 0212: tokens predicted by the LLM are organized into one or more sequences, each sequence forming a distinct synthetic query; Paragraph 0039: “all or aspects of the NL based response system 120 can be implemented locally at the client device 110,” wherein the client device locally stores and executes the LLM performing the query generation.)
Therefore, it would have been obvious to one of ordinary skill in the art prior to the effective filing date of the invention to modify Ellis's chat-service search term extraction to be performed by a Large Language Model stored on the terminal apparatus, as taught by Rofouei. One would have been motivated to do this in order to generate search queries that account for the full conversational context of the chat session rather than isolated keyword matches against a stored dictionary, thereby mitigating over-specified or under-specified queries and returning promoted content that better resonates with the user's ongoing conversation.
The Ellis/Rofouei combination does not appear to explicitly disclose that the LLM is trained by a search query that was input... in the past when conversion was achieved with respect to an advertisement displayed in search advertising, such that the extraction yields a search query by which an advertisement that is highly likely to achieve conversion is displayed.
Alakuijala, however, discloses a generative query model that is trained by a search query that was input... in the past when conversion was achieved with respect to an advertisement displayed in search advertising, wherein the model generates a search query by which an advertisement that is highly likely to achieve conversion is displayed (Paragraph 0011: a generative model is trained on training data based on past query submissions identified based on users selecting shopping content, e.g., certain ads, in association with the submissions, and based on users completing a transaction following the submissions; Paragraph 0014: advertisements or other content are provided to the client device that generated the query based on such content being assigned to one or more of the variants generated by the generative model; Paragraphs 0036 and 0039: one or more aspects of the systems and engines, including the variant engine operating over the trained generative models, may be implemented on the client device 106.)
Therefore, it would have been obvious to one of ordinary skill in the art prior to the effective filing date of the invention to train the LLM of the Ellis/Rofouei combination using past search queries identified based on users selecting advertisements and completing transactions following the queries, as taught by Alakuijala, such that the queries extracted from the chat conversation are queries for which displayed advertisements are likely to achieve conversion. One would have been motivated to do this in order to bias the query generation toward queries with demonstrated commercial effectiveness, thereby increasing the likelihood that the promotion links returned for the extracted query result in completed transactions.
The Ellis/Rofouei/Alakuijala combination does not appear to explicitly disclose that the training search query was input by the user whose conversation information is being processed.
Abdallah, however, discloses a model trained by data that was input by the user in the past, wherein the trained model is applied to the same user's subsequent conversation (col. 4, ll. 55-65: “The conversation memory module 142 is configured to supply a long term memory of aspects of categories as searched by a particular user using machine learning”; col. 5, ll. 44-60: data from a first natural-language conversation of the user is used to train a model, and a second natural-language conversation of the user is then processed using the model to automatically include aspects from the prior conversation.)
Therefore, it would have been obvious to one of ordinary skill in the art prior to the effective filing date of the invention to train the model of the Ellis/Rofouei/Alakuijala combination on search queries previously input by the same user whose conversation information is being processed, as taught by Abdallah. One would have been motivated to do this in order to personalize the query extraction to the individual user's demonstrated purchasing behavior and reduce repetitive user input, consistent with Abdallah's stated objective of improving the operational efficiency of computing devices employing natural language conversations.
Claim 2: The Ellis/Rofouei/Alakuijala/Abdallah combination discloses those limitations cited above.
Ellis further discloses wherein the processor is further configured to make a request to the information processing apparatus for an advertisement content based on the search query when the user selects the advertisement link, and wherein the processor displays the advertisement content that is transmitted from the information processing apparatus in accordance with the request (Paragraph 0112: “links 714 that upon selection cause the consumer interface to enter an electronic purchase funnel, or the like, for the linked promotion”; Paragraph 0063: “The server 104 may facilitate the generation and providing of various electronic marketing communications, including output chat data for mobile chat sessions, based on input chat data received from the mobile chat session.”)
Claim 9: The Ellis/Rofouei/Alakuijala/Abdallah combination discloses those limitations cited above.
Abdallah further discloses wherein the processor uses a model that is trained to output the search query while adopting the conversation information as input, and adopts the search query that is output from the model as an extraction result (Abstract: “data that describes this first natural-language conversation is used to train a model using machine learning . . . processed using the model as part of machine learning to generate the second search query”; col. 5, ll. 44-60: data from a first natural-language conversation is processed using a trained model to automatically generate a search query as output without user intervention; Fig. 3, machine learning module 128, model 316, search query 326, depicting the pipeline from conversation input data through the trained model to search query output.)
The rationale for combining Abdallah with Ellis, Rofouei, and Alakuijala is articulated above and reincorporated herein by reference.
Claims 3-4 and 6-8 are rejected under 35 U.S.C. § 103 as being unpatentable over Ellis/Rofouei/Alakuijala/Abdallah in view of Warren (US 11,070,498 B2).
Claim 3: The Ellis/Rofouei/Alakuijala/Abdallah combination discloses those limitations cited above, but does not appear to explicitly disclose wherein the processor displays a search result content that includes the advertisement content and a search result based on the search query.
Warren, however, discloses wherein the processor displays a search result content that includes the advertisement content and a search result based on the search query (col. 9, ll. 42-55: “In the search results page (see FIG. 4, item 410) returned to the user, a list of related search queries recently submitted to the search engine by other users is presented to the user (see FIG. 4, item 412)”; Figs. 4, 13, 14, wherein both sponsored conversation links, i.e., advertisement content, and the underlying web search results based on the user's search query are displayed within a single unified results page.)
Therefore, it would have been obvious to one of ordinary skill in the art prior to the effective filing date of the invention to combine this feature of Warren with those of the Ellis/Rofouei/Alakuijala/Abdallah combination. One would have been motivated to do this in order to provide users with a unified, contextually relevant interface that increases engagement with both the advertisement content and the underlying search results.
Claim 4: The Ellis/Rofouei/Alakuijala/Abdallah/Warren combination discloses those limitations cited above.
Abdallah further discloses wherein the search result is a search result that was retrieved based on the search query in the past by the user (col. 4, ll. 55-65: “The conversation memory module 142 is configured to supply a long term memory of aspects of categories as searched by a particular user using machine learning”; col. 5, ll. 44-60: data from a first natural-language conversation is used to train a model, and a second natural-language conversation is then processed using the model to automatically include aspects from the prior conversation, wherein search aspects from a user's past search query are retained and reused to generate results in a subsequent session.)
The rationale for combining Abdallah with Ellis, Rofouei, and Alakuijala is articulated above and reincorporated herein by reference.
Claim 6: The Ellis/Rofouei/Alakuijala/Abdallah/Warren combination discloses those limitations cited above.
Warren further discloses wherein the processor displays the search result content at a timing that is designated by the user (claim 3: “The computer-implemented method as claimed in claim 1 wherein the link is only generated for the first user if the sponsored concept and the first topic of interest are received within a user-configurable time period”; col. 7, ll. 57-65: “User A can then type a message into transmitted text box 510 and click the Send button”; claim 10: “the agent of the sponsoring company or the first user may decline the invitation to converse”, wherein the user controls the timing at which the linked advertisement and search result content is displayed through user-configurable time settings and user-activated link selection.)
The rationale for combining Warren with Ellis, Rofouei, Alakuijala, and Abdallah is articulated above and reincorporated herein by reference.
Claim 7: The Ellis/Rofouei/Alakuijala/Abdallah/Warren combination discloses those limitations cited above.
Ellis further discloses wherein the processor shares the search result content with a conversation partner of the user in the chat service (Paragraphs 0130-0138: disclosing receipt and processing of chat data from both a first and second consumer device connected to the same mobile chat session; Fig. 9, element 916, showing the promotion links displayed within message display 902 of the shared chat session visible to all participants.)
Claim 8: The Ellis/Rofouei/Alakuijala/Abdallah/Warren combination discloses those limitations cited above.
Warren further discloses wherein the processor shares the search result content with the conversation partner when the user gives a permission (Fig. 6, alert dialog 610, depicting the prompt “Do you accept? OK Cancel” presented to User B before any content sharing commences; col. 7, l. 57 through col. 8, l. 20: User B receives the alert dialog 610 and content sharing between the users is initiated only upon User B's affirmative acceptance of the invitation by clicking a button; claim 10: “the agent of the sponsoring company or the first user may decline the invitation to converse”, wherein the sharing of search result content with the conversation partner is expressly conditioned on the user's affirmative grant of permission via the accept/decline dialog.)
The rationale for combining Warren with Ellis, Rofouei, Alakuijala, and Abdallah is articulated above and reincorporated herein by reference.
Claim 5 is rejected under 35 U.S.C. § 103 as being unpatentable over Ellis/Rofouei/Alakuijala/Abdallah/Warren in view of Garrett et al. (US 10,621,243 B1).
The Ellis/Rofouei/Alakuijala/Abdallah/Warren combination discloses those limitations cited above, but does not appear to explicitly disclose wherein the search result is a search result that is retrieved based on the search query when the information processing apparatus receives a request for the advertisement content from the processor.
Garrett, however, discloses wherein the search result is a search result that is retrieved based on the search query when the information processing apparatus receives a request for the advertisement content from the processor (col. 3, ll. 13-22: “automatically submitting to a search engine a search query received in association with the reserved term, if the user input contains a reserved term, receiving one or more search results from the search engine, and automatically providing, in the electronic communication channel . . . at least one of the one or more search results”; Fig. 3A, steps 1-6, illustrating the real-time sequence of monitoring the conversation, receiving the query, submitting it to the search engine, receiving the response, and inserting it into the conversation, wherein the search result retrieval occurs contemporaneously with, and is triggered by, the request for the search service.)
Therefore, it would have been obvious to one of ordinary skill in the art prior to the effective filing date of the invention to combine this feature of Garrett with those of the Ellis/Rofouei/Alakuijala/Abdallah/Warren combination. One would have been motivated to do this in order to ensure that the search results delivered to the user are current and responsive to the user's immediate informational need at the moment of the request.
Claim 12 is rejected under 35 U.S.C. § 103 as being unpatentable over Ellis/Rofouei/Alakuijala/Abdallah in view of Lyren (US 2014/0330649 A1).
The Ellis/Rofouei/Alakuijala/Abdallah combination discloses those limitations cited above. Ellis further discloses wherein the processor displays the advertisement link in a talk room for a conversation partner in the chat service, and wherein the advertisement link is displayed as utterance of a third person other than the user and the conversation partner (Paragraph 0042: the chat service is “configured to interface with mobile chat applications to provide electronic marketing communications as chat messages”; Paragraphs 0130-0138: disclosing a mobile chat session to which both a first and second consumer device are connected; Fig. 9, element 916, showing the promotion links inserted into the message display 902 of the shared chat session as chat messages originating from the promotion service rather than from either chat participant.)
The Ellis/Rofouei/Alakuijala/Abdallah combination does not appear to explicitly disclose wherein the advertisement link is displayed for only the user between the user and the conversation partner.
Lyren, however, discloses wherein the advertisement link is displayed for only the user between the user and the conversation partner (Paragraph 0058: during a text exchange between two users, a targeted video advertisement “displays on the HPED of the Ben, but not on the HPED of the Jerry”; Paragraph 0050: while two users are talking to each other, an advertisement plays on the display of the electronic device of the first user while a different advertisement simultaneously plays on the display of the electronic device of the second user, and “The first user does not see the advertisement for the smartphone, and the second user does not see the advertisement for the tennis shoes”; Paragraph 0046: during a video call, the targeted advertisements are displayed such that each advertisement appears on only one participant's display.)
Therefore, it would have been obvious to one of ordinary skill in the art prior to the effective filing date of the invention to display the advertisement chat message of the Ellis/Rofouei/Alakuijala/Abdallah combination on only the terminal apparatus of the user, and not on the device of the conversation partner, as taught by Lyren. One would have been motivated to do this in order to deliver an advertisement targeted to one participant's interests without exposing that participant's targeted content to the other party to the conversation.
Claim 13 is rejected under 35 U.S.C. § 103 as being unpatentable over Ellis/Rofouei/Alakuijala/Abdallah in view of Makar et al. (US 2012/0095835 A1).
The Ellis/Rofouei/Alakuijala/Abdallah combination discloses those limitations cited above, but does not appear to explicitly disclose wherein the processor displays the advertisement link at a timing at which a conversation between the user and a conversation partner is paused, wherein the timing is a timing after a lapse of a predetermined time or more since last utterance.
Makar, however, discloses displaying promotional chat content at a timing... after a lapse of a predetermined time or more since last utterance (Paragraph 0019: “the present invention launches a chat window during a session, such as after a settable period of time, or after settable period of user inactivity or a combination of both”; Paragraph 0074: launching of the chat window is triggered by activities including “no activity from the end-user within a settable period of time” and an “inactivity timer expiring”; Paragraph 0054: upon expiration of the delay timer, the sales pitch is printed in the chat window.)
Therefore, it would have been obvious to one of ordinary skill in the art prior to the effective filing date of the invention to display the advertisement chat message of the Ellis/Rofouei/Alakuijala/Abdallah combination upon the lapse of a settable period of inactivity since the last message in the chat session, as taught by Makar. One would have been motivated to do this in order to present the promotional content at a moment of user inactivity, thereby re-engaging the user during a lull rather than interrupting an active exchange.
Claim 14 is rejected under 35 U.S.C. § 103 as being unpatentable over Ellis/Rofouei/Alakuijala/Abdallah in view of Luna (US 9,407,713 B2).
The Ellis/Rofouei/Alakuijala/Abdallah combination discloses those limitations cited above, but does not appear to explicitly disclose wherein the processor stores therein a search result that was displayed in response to input of the search query by the user in the past, and makes a request to an information processing apparatus for only an advertisement that is displayed based on the search query.
Luna, however, discloses wherein the processor stores therein a search result that was displayed in response to input of the search query by the user in the past, and makes a request to a server for only content not available from the stored results (col. 15, ll. 48-51: the caching policy manager locally stores responses for data requests in the local database as cache entries for subsequent use in satisfying same or similar data requests; col. 15, ll. 19-26: upon intercepting a data request, the local repository is queried to determine if a locally stored response is available and valid, and when a valid response is available in the local cache, the response is provided to the application without the device needing to access the cellular network; col. 15, ll. 36-43: only if a valid cache response is not available is the data request sent onward to the remote content source; col. 8, ll. 53-57: serving requests from the local cache reduces the number of requests that need to be satisfied over the network; claims 1-2.)
Therefore, it would have been obvious to one of ordinary skill in the art prior to the effective filing date of the invention to locally store, on the terminal apparatus of the Ellis/Rofouei/Alakuijala/Abdallah combination, the search result previously displayed for the user's search query, and to request from the information processing apparatus only the advertisement content that is not satisfiable from the local storage, as taught by Luna. One would have been motivated to do this in order to reduce the number of requests transmitted over the network, thereby conserving network bandwidth and device battery consumption while still obtaining the current advertisement content.
Other Relevant Prior Art
Though not cited in the aforementioned rejections, the following references are nevertheless deemed to be relevant to Applicant’s disclosures:
Clark et al. (10296949), directed to a messenger application plug-in for providing tailored ads within a conversation thread.
Buchheit et al. (20130013634), directed to retrieving conversations that match a search query.
Rathod et al. (20220179665), directed to displaying user related contextual keywords and controls for user selection and storing and associating selected keywords and user interaction with controls data with user.
Buchheit et al. (8601062), directed to providing snippets relevant to a search query in a conversation-based email system.
Response to Arguments
Claim Interpretation Under 35 U.S.C. § 112(f)
Applicant's arguments regarding the interpretation of “extraction unit,” “generation unit,” “display processing unit,” and “requesting unit” under 35 U.S.C. § 112(f) are persuasive. In view of the amendment replacing these terms with “a processor configured to” perform the recited functions, the claims now recite sufficient structure, and the § 112(f) interpretation is withdrawn.
Rejections Under 35 U.S.C. § 101
Applicant's arguments with respect to the rejection of the claims under 35 U.S.C. § 101 have been fully considered but are not persuasive.
Applicant first argues, citing Ex parte Desjardins, Appeal No. 2024-000567 (PTAB Sept. 26, 2025), that the claims integrate any abstract idea into a practical application because the specification describes that training the LLM on conversion-achieving search queries enables extraction of conversion-optimized queries, which Applicant characterizes as an improvement to advertisement distribution efficiency (Specification, pages 4-5, 9). Examiner respectfully disagrees. The case is distinguishable on its facts: in Desjardins, the specification identified improvements to how the machine learning model itself operates, and the claims reflected those model-level improvements. Here, by contrast, the claims recite a generically-claimed “Large Language Model (LLM)” and do not purport to improve the operation, architecture, or training mechanics of the LLM itself. The claims instead select a particular corpus of training data (past conversion-achieving search queries) and apply the generically-recited model to a particular data environment (chat conversation information) for a particular commercial end (displaying advertisements likely to achieve conversion). This is precisely the circumstance addressed in Recentive Analytics, Inc. v. Fox Corp., No. 2023-2437 (Fed. Cir. Apr. 18, 2025): the application of generic machine learning to a new data environment, without improvement to the machine learning models themselves, does not confer eligibility. Moreover, the asserted improvement to “advertisement distribution efficiency” (Specification, page 9) is an improvement to the abstract idea of targeted advertising itself, not an improvement to the functioning of a computer or to any other technology or technical field. See MPEP § 2106.05(a) (the alleged improvement must be to technology, and an improvement to the judicial exception itself is not sufficient); MPEP § 2106.04(d). A more effective scheme for selecting which advertisements to show remains an advertising scheme.
Applicant next argues that the claims recite a specific technical architecture because the processor of the terminal apparatus itself stores and uses the trained LLM. This argument is likewise not persuasive. The terminal apparatus and its processor are recited at a high level of generality, and the claim merely specifies where generically-recited computer components perform the abstract data analysis. Locating generic storage and generic processing at the terminal rather than at a server amounts to an instruction to implement the abstract idea on a particular generic computer, which is insufficient under MPEP § 2106.05(b) and § 2106.05(f). See Alice Corp. Pty. Ltd. v. CLS Bank Int'l, 573 U.S. 208, 223 (2014).
Applicant further argues that new claims 12 and 13 provide technical specificity reinforcing eligibility, characterizing claim 12 as a privacy-preserving display mechanism not performable by a human with pen and paper and claim 13 as a specific technical trigger for non-intrusive display. Examiner respectfully disagrees. Claim 12 specifies to whom the advertisement is shown and in what guise (as an utterance of a third person, visible only to the user), and claim 13 specifies when the advertisement is shown (after a lapse of a predetermined time since the last utterance). These limitations define an advertising presentation strategy and timing parameters that further characterize the abstract idea of targeted advertising; they do not improve the functioning of the computer or any other technology. See MPEP § 2106.04(a)(2). The Prong One analysis, moreover, rests on the certain-methods-of-organizing-human-activity grouping in addition to the mental-process grouping, and the commercial character of choosing how and when to present an advertisement to a customer is not diminished by generic computerized display and timing functions. The display acts themselves are well-understood, routine, and conventional computer functions as set forth in the Berkheimer analysis above.
For at least these reasons, the rejection of claims 1-14 under 35 U.S.C. § 101 is maintained.
Rejections Under 35 U.S.C. § 102
Applicant's arguments with respect to the rejection of claims 1-2, 7, and 10-11 under 35 U.S.C. § 102(a)(2) over Ellis are moot in view of the new grounds of rejection under 35 U.S.C. § 103 set forth above, which were necessitated by Applicant's amendment.
Rejections Under 35 U.S.C. § 103
Applicant's arguments with respect to the rejections of claims 3-6 and 8 under 35 U.S.C. § 103, insofar as they are directed to the amended features of independent claim 1, are moot in view of the new grounds of rejection set forth above, which were necessitated by Applicant's amendment.
Regarding claim 9, Applicant argues that Abdallah's machine learning model is trained to remember user-specified product aspects across conversations (Abdallah, col. 6, lines 47-60), not to extract search queries based on past conversion-achieving search behavior in search advertising. Examiner agrees that Abdallah is not relied upon for training based on conversion-achieving search behavior. That feature is taught by Alakuijala, which discloses training the generative query model on past query submissions identified based on users selecting advertisements in association with the submissions and completing transactions following the submissions (Alakuijala, Paragraph 0011), as set forth in the rejection above. Abdallah is relied upon for the distinct teaching that the training data is drawn from the same user whose subsequent conversation is processed (Abdallah, col. 4, lines 55-65; col. 5, lines 44-60), and for the model-based extraction recited in claim 9. The rejections are based on the combined teachings of the references, and nonobviousness cannot be established by attacking references individually where the rejection is based on a combination. See MPEP § 2145(IV); In re Keller, 642 F.2d 413, 208 USPQ 871 (CCPA 1981).
Regarding Applicant's assertion that none of the previously cited references disclose the split retrieval architecture of new claim 14, this limitation is addressed by Luna as set forth in the rejection of claim 14 above.
New Claims 12-14
New claims 12-14 are rejected as set forth in detail above. Applicant's arguments that claims 12-13 recite additional technical elements reinforcing eligibility are addressed in the § 101 response above.
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 CHRISTOPHER BUSCH whose telephone number is (571)270-7953. The examiner can normally be reached M-F 10-7.
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, Waseem Ashraf can be reached at 571-270-3948. 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.
/CHRISTOPHER C BUSCH/Examiner, Art Unit 3621
/WASEEM ASHRAF/Supervisory Patent Examiner, Art Unit 3621
Prosecution Insights