DETAILED ACTION
This office action is responsive to communication(s) filed on 6/10/2026.
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 .
Claims Status
Claims 1-20 are pending and are currently being examined.
Claims 1, 10 and 16 are independent and are newly amended.
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 of this title, 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) 1-2, 4-7, 9-13, 15-17 and 19-20 is/are rejected under 35 U.S.C. 103 as being unpatentable over Guinn; Devon et al. (hereinafter Guinn – US 20240412030 A1) in view of Langdon, JR.; Charles A. (hereinafter Langdon – US 20210295434 A1), Lee; Ji-yeon et al. (hereinafter Lee – US 20180150905 A1), Chica; Sebastian de la et al. (hereinafter Chica -- US 20130060763 A1) and Tian; Wei (hereinafter Tian – US 20160188146 A1).
Independent Claim 1:
Guinn teaches:
A computer-implemented method for generating a customized output (computer system/method, Abstract and fig. 1)
based on account data of a user, (e.g., user identifier, ¶ 119, personas, embeddings, memories, ¶¶ 81 and 130)
the computer-implemented method comprising:
receiving, by one or more processors, user data (during a training phase, input information of related user is received, e.g., user’s favorites/preferences and converted into embeddings, ¶¶ 106-107 and figs. 6, and these embeddings are stored in association with the user, Abstract and ¶¶ 57-59, e.g., in a knowledge layer, ¶ 100. Using such inputs, users create personalized AI personas by employing natural language, documents, and UI-based customization to define specific knowledge, memories, and communication styles, which are then used to perform targeted tasks for the user or to interact with a broader audience, Abstract and ¶ 81. Note that Guinn makes references a “first” and “second individual”, a “second individual” can refer to the same or different person as a “first individual”, ¶ 5)
wherein the user data includes user browser history, (user related information includes, e.g., memories including query history, ¶ 46, wherein the queries are in a Web Browser environment [user browser history], ¶ 55. )
user literacy level, (Paragraph 78 describes using digital avatars for tailored education, which implies adapting content to user literacy and comprehension levels through "personalized educational experiences" (adaptive learning). This is supported by ¶ 33 of Guinn, which teaches that neural network outputs are customized based on an individual's knowledge, directly connecting, in this context, to user literacy, ¶¶ 33 and 78.)
and user account data, (e.g., user identifier, ¶ 119, embeddings/memories, ¶ 130)
[…];
creating, the one or more processors, a user record based on the user account data, (a persona is created by a training engine [withing a computer system comprising one or more processors] for a user, ¶¶ 37, 39 and 52. This paragraph reflects a machine learning infrastructure that is creating a user record because it describes a training engine that is collecting user-provided or selected information to update a classifier or regression model, which in turn acts as a personalized, stored "persona" [or user record]. Information of the user may be stored, e.g., in association with an embedding [user record] data structure, ¶¶ 118-119)
[…];
retrieving, by the one or more processors, one or more relevant segments from a vector database, (embeddings of the user related information are stored as portion(s) [segments] of vectorized knowledge and used to match, access, search, and use the vast amount of data in the knowledge layer [retrieving], ¶¶ 102 and 130)
wherein the one or more relevant segments correspond to one or more relevant events (the relevant segments may be used for representing and matching certain topics [correspond to one or more relevant events] about the user, e.g., favorite foods, preferences, etc., ¶¶ 107 and 133)
that are relevant to the [financial investments], (the system offers personalized financial advice and investment recommendations, ¶ 79)
[…];
generating, by a machine learning model, a personalized output script based on the one or more relevant segments (as noted above, the system offers “personalized” financial advice and investment recommendations, ¶ 79. A user can train a persona by creating custom scripts to “ensuring that each interaction is unique and tailored to the needs of the users' audience” [personalized output script], ¶ 86. A computation engine can automatically perform reinforced learning or generate a revised classier or regression model, ¶ 53. As such, this is reflective of an automatic reinforcement training of a digital human avatar, combined with user-created custom scripts, and is considered a form of generating or dynamically refining scripts [personalized output script]. In this context, the training process uses user input to shape the AI's behavior, allowing the avatar to generate unique, context-aware responses or scripts in real-time. Because the computation engine is modifying the training/persona, it is considered part of a machine learning model [generating, by a machine learning model, a personalized output script])
and the user literacy level, […], (As explained above, in reference to ¶¶ 33 and 78, Guinn teaches that neural network outputs are customized based on an individual's knowledge, directly connecting, in this context, to user literacy, ¶¶ 33 and 78.)
wherein the machine learning model modifies the personalized output script based on the user literacy level; (As reflected in the above mapping, the machine learning model dynamically updates and refines personalized educational and financial scripts through reinforcement learning based on individual user knowledge/literacy level. By analyzing user input and utilizing neural networks to tailor content, the system, as supported by ¶¶ 33, 53, 78-79, and 86, continuously adapts its output to match the specific comprehension level of the audience)
creating, by the machine learning model, an output based on the personalized output script, (a personalized response [generating…a personalized output script] is generated [creating…an output based on the personalized output script] by the neural network [machine learning model], ¶¶ 138-139 and fig. 9:938)
the output including one or more graphic visuals corresponding to the personalized output script; (the output may be in the form of displayable text [one or more graphic visuals], fig. 9:946 and ¶¶ 139-140)
and displaying, by the one or more processors, the output on one or more user interfaces of a user device. (dialogue output, e.g., displayed via text, fig. 9:94-946 and ¶¶ 139-140)
Guinn further suggests:
that the user data is received from one or more databases, (Guinn teaches the concept of retrieving information from database(s), e.g., vector database, ¶ 118, and that the data can be pulled from external sources, for extending the versatility and reach of the method, ¶ 88)
Accordingly, it would have been obvious to a person having ordinary skill in the art, before the effective filing date of the claimed invention, to modify the method of Guinn to include that the user data is received from one or more database, as suggested by Guinn.
One would have been motivated to make such a combination in order to improve the versatility and reach of the method by obtaining information from external databases, Guin ¶¶ 88 and 118.
Guinn does not appear to expressly teach, but Langdon teaches:
wherein the user record includes one or more stocks; (Searches, include information such “stock symbols” [one or more stocks], Langdon ¶ 125)
determining, by the one or more processors, a stock subset of the one or more stocks; (e.g., filter by top ranked companies, ¶ 138)
that the financial investment are the stock subset (e.g., filter by top ranked companies, ¶ 138)
Accordingly, it would have been obvious to a person having ordinary skill in the art, before the effective filing date of the claimed invention, to modify the method of Guinn to include wherein the user record includes one or more stocks, determining, by the one or more processors, a stock subset of the one or more stocks, and that the financial investment are the stock subset, as taught by Langdon.
One would have been motivated to make such a combination in order to improve the versatility of the method by modifying the method to apply to financial stock investment advice in a fast and functional way, Langdon 26 and Guinn ¶ 79.
Guinn, as modified, does not appear to expressly teach, but Lee teaches:
the personalized output script comprising a summarization of content of the one or more relevant segments (a machine learning unit that is able to generate summarized content related to one or more segments, e.g., summarization range selection, Abstract and ¶¶ 3, 21 and 57, and figs. 5A-5C).
Accordingly, it would have been obvious to a person having ordinary skill in the art, before the effective filing date of the claimed invention, to further modify the method of Guinn to include the personalized output script comprising a summarization of content of the one or more relevant segments, as taught by Lee.
One would have been motivated to make such a combination in order to improve the efficiency and convenience offered by the method, e.g., by provide the user the ability to check/understand “a large amount of content more quickly and conveniently”, Lee ¶ 5.
Guinn further teaches that the user may browse a web page or website, ¶ 93.
Guinn, as modified, does not appear to expressly teach, but Chica teaches:
wherein the user literacy level is determined by analyzing the user browser history including search history and types of [documents] visited by the user (A system infers a user's reading [literacy] level for personalized search results by analyzing their browsing behavior, including past search queries [search history], long-term browsing history, and the reading difficulty of previously visited documents [this reading difficulty of the documents serves as a type of articles and web pages visited by the user], ¶¶ 12-13. Furthermore, this process leverages a document reading difficulty modeler to assess content difficulty, along with a machine learning model, such as LambdaMART, to rank search results based on user-specific reading level profiles, ¶¶ 30 and 50. It was well within the capabilities of a person having ordinary skill in the art to have realized that the documents browsed in the Internet is often implemented as articles and web pages).
Accordingly, it would have been obvious to a person having ordinary skill in the art, before the effective filing date of the claimed invention, to further modify the method of Guinn to include wherein the user literacy level is determined by analyzing the user browser history including search history and types of [documents] visited by the user, as taught by Chica.
One would have been motivated to make such a combination in order to improve the method by provided personalized browsing of web page documents to the users based on user activity, Chica ¶ 14 and Guinn ¶ 93.
Guinn, as modified, does not appear to expressly teach, but Tian teaches:
that the documents are articles (typical internet documents include webpages and articles, e.g., webpage content including a “news article webpage” as standard content users navigate while browsing, ¶ 24).
Accordingly, it would have been obvious to a person having ordinary skill in the art, before the effective filing date of the claimed invention, to further modify the method of Guinn to include that the documents are articles, taught by Tian.
One would have been motivated to make such a combination in order to improve the practicality of the method by browsing common types of web page documents, Tian ¶ 24.
Claim 2:
The rejection of claim 1 is incorporated. Guinn further teaches:
the computer-implemented method further comprising:
storing, by the one or more processors, the output in a database record in the one or more databases, (the data structure or database return relevant memories or knowledge to the correct user, ¶ 130, which includes prior conversations, ¶ 66)
wherein the database record includes a reference to the user record. (user identifier used to match to correct user, ¶ 130)
Claim 4:
The rejection of claim 1 is incorporated. Guinn further teaches:
wherein the output includes at least one of:
a video, a text-based output, or an audio output. (the output may include text, audio, and/or an image, ¶ 8)
Claim 5:
The rejection of claim 1 is incorporated. Guinn, as modified, Langdon further teaches:
the computer-implemented method further comprising:
ranking, by the one or more processors, the one or more stocks of the user record based on a priority, wherein the priority is based on one or more heuristic rules. (Guinn teaches that the computational techniques include machine learning techniques [based on one or more heuristic rules], Langdon teaches ranking the investment data, ¶¶ 80 and 100, e.g., by top ranked companies, based on certain criteria ¶ 138, and rules, e.g., machine learning logic [based on one or more heuristic rules], ¶ 80)
Claim 6:
The rejection of claim 5 is incorporated. Guinn, as modified, Langdon further teaches:
wherein the ranking is performed by a machine-learning model. (Langdon teaches ranking the investment data, ¶¶ 80 and 100, e.g., by top ranked companies, based on certain criteria ¶ 138, and rules, e.g., machine learning logic [based on one or more heuristic rules], ¶ 80)
Claim 7:
The rejection of claim 1 is incorporated. Guinn further teaches:
wherein retrieving the one or more relevant segments from the vector database comprises:
building, by the one or more processors, the vector database; (converting entries to vector-based embeddings, e.g., see ¶¶ 107 and 126)
and querying, by the one or more processors, the vector database for the one or more relevant events and the corresponding one or more relevant segments. (searching [querying] the embeddings for matching certain criteria [relevant events] and user identifier [relevant segments], ¶¶ 102)
Claim 9:
The rejection of claim 7 is incorporated. Guinn, as modified, further teaches:
wherein querying the vector database comprises:
receiving, by the one or more processors, the stock subset; (Guinn teaches information of related to user is received, e.g., user’s favorites/preferences, ¶ 107 and figs. 6, and stored in association with the user, Abstract and ¶¶ 57-59. Langdon teaches that the information includes one or more stocks, Langdon ¶ 125)
processing, by the one or more processors, stock information corresponding to each of the one or more stocks in the stock subset to produce one or more vector queries, (the vectorized information is used for searching/querying, Guinn ¶¶ 44 and 102. Langdon teaches that the information includes one or more stocks, Langdon ¶ 125)
the stock information including at least one stock ticker, at least one name, or at least one company name; (stock ticker, or company name, Langdon ¶ 151)
and querying, by the one or more processors, the vector database for information related to each of the one or more vector queries. (the vectorized information is used for searching/querying, Guinn ¶¶ 44 and 102, and is stored in a database, Guinn ¶ 118)
Independent Claims 10 and 16:
Claim(s) 10 and 16 is/are directed to a system and a medium for accomplishing the steps of the method in claim 1, and are rejected using similar rationale(s).
Claims 11 and 20:
The rejection of claims 10 and 16 are incorporated. Claim(s) 11 and 20 is/are directed to a system and a medium for accomplishing the steps of the method in claim 4, and are rejected using similar rationale(s).
Claim 12:
The rejection of claim 10 is incorporated. Claim(s) 12 is/are directed to a system for accomplishing the steps of the method in claim 5, and is rejected using similar rationale(s).
Claim 13:
The rejection of claim 12 is incorporated. Claim(s) 13 is/are directed to a system for accomplishing the steps of the method in claim 6, and is rejected using similar rationale(s). However, note the correction required, as explained in 112(b) rejection section above, due to insufficient antecedent basis.
Claims 15 and 17:
The rejection of claim 10 and 16 are incorporated. Claim(s) 15 and 17 is/are directed to a system and a medium for accomplishing the steps of the method in claim 7, and are rejected using similar rationale(s).
Claim 19:
The rejection of claim 17 is incorporated. Claim(s) 19 is/are directed to a medium for accomplishing the steps of the method in claim 9, and are rejected using similar rationale(s).
Claim(s) 3 and 14 is/are rejected under 35 U.S.C. 103 as being unpatentable over Guinn (US 20240412030 A1) in view of Langdon (US 20210295434 A1), Lee (US 20180150905 A1), Chica (US 20130060763 A1) and Tian (US 20160188146 A1), as applied to claims 1 and 10 above, and further in view of Luo; Ping et al. (hereinafter Luo – US 20150127602 A1).
Claim 3:
The rejection of claim 1 is incorporated. Guinn teaches that customization of output may be based on a second individual different from a first, Abstract.
Guinn, as modified, does not appear to expressly teach, but Luo teaches:
wherein the one or more stocks include at least one of:
at least one stock owned by the user, at least one stock previously owned by the user, at least one stock followed by the user, at least one stock owned by one or more other users that the user follows, at least one stock watched by the one or more other users that the user follows, at least one stock owned by one or more other users with a similar portfolio, or at least one stock watched by the one or more other users with the similar portfolio. (an investment portfolio recommendation application that mines the data of experienced investors to identify quality or interesting patterns and make investment plan recommends to new investors based on these patterns, ¶ 14).
Accordingly, it would have been obvious to a person having ordinary skill in the art, before the effective filing date of the claimed invention, to further modify the method of Guinn to include wherein the one or more stocks include at least one of: at least one stock owned by the user, at least one stock previously owned by the user, at least one stock followed by the user, at least one stock owned by one or more other users that the user follows, at least one stock watched by the one or more other users that the user follows, at least one stock owned by one or more other users with a similar portfolio, or at least one stock watched by the one or more other users with the similar portfolio, as taught by Luo.
One would have been motivated to make such a combination in order to improve the quality of investment recommendations provided by the method, Luo ¶ 14.
Claim 14:
The rejection of claim 10 is incorporated. Claim(s) 14 is/are directed to a system for accomplishing the steps of the method in claim 3, and is rejected using similar rationale(s).
Claim(s) 8 and 18 is/are rejected under 35 U.S.C. 103 as being unpatentable over Guinn (US 20240412030 A1) in view of Langdon (US 20210295434 A1) and Lee; Ji-yeon et al. (hereinafter Lee – US 20180150905 A1), Chica (US 20130060763 A1) and Tian (US 20160188146 A1), as applied to claims 7 and 17 above, and further in view of Smith Lewis; Andrew et al. (hereinafter Smith – US 20240289863 A1).
Claim 8:
The rejection of claim 7 is incorporated. Although the added limitations of this claim seem to be common steps for building and maintaining a vector database, assuming arguendo that that they are not, Guinn, as modified, does not appear to expressly teach, but Smith teaches:
wherein the building the vector database comprises:
for one or more time periods, analyzing, by the one or more processors, one or more sources of information; (e.g., user profile information is used for creating vectors, ¶ 72, and the profile is continually and periodically updated [analyzing] to ensure accuracy and personalization of system outputs, ¶ 116)
converting, by the one or more processors, data of the one or more sources to one or more short segments using a machine-learning model; (splitting documents into smaller chunks, ¶ 46, machine learning usable for the processing, ¶ 60)
storing, by the one or more processors, the one or more short segments in the one or more databases; (storing the chunks, ¶ 46, e.g., in a vector database, ¶ 49)
processing, by the one or more processors, the one or more short segments to determine a corresponding latent vector for each of the one or more short segments; (chunks are pass through an embedding model to generate vectors [latent vector or embedding], ¶ 50)
and storing, by the one or more processors, the processed one or more short segments and the corresponding latent vector in the vector database (vectors and chunks are stored in vector database, ¶ 49).
Accordingly, it would have been obvious to a person having ordinary skill in the art, before the effective filing date of the claimed invention, to further modify the method of Guinn to include wherein the building the vector database comprises: for one or more time periods, analyzing, by the one or more processors, one or more sources of information; converting, by the one or more processors, data of the one or more sources to one or more short segments using a machine-learning model; storing, by the one or more processors, the one or more short segments in the one or more databases; processing, by the one or more processors, the one or more short segments to determine a corresponding latent vector for each of the one or more short segments; and storing, by the one or more processors, the processed one or more short segments and the corresponding latent vector in the vector database, as taught by Smith.
One would have been motivated to make such a combination in order to ensure the accuracy and personalization of outputs produced by applying the method, Smith ¶ 116.
Claim 18:
The rejection of claim 17 is incorporated. Claim(s) 18 is/are directed to a medium for accomplishing the steps of the method in claim 8, and is rejected using similar rationale(s).
Response to Arguments
The previous 101 rejections have been overcome by claim amendments. One or more of the applicant’s 101 arguments directed to a practical application, see Remarks Pg(s) are partially persuasive. The practical application is that the purported improvement in efficiency and overall user experience in the retrieval/presentation efficiency of financial information, see Instant Specification (as published) ¶¶ 2 and 14, is reflected in the amended claims.
The previous 112(a) and 112(b) rejections have been overcome by claim amendment.
Applicant's 103 arguments have been fully considered but they are moot in view of the new grounds of rejection presented 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.
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure. Below is a list of these references, including why they are pertinent:
Wang; Jun et al. US 20210073237 A1, is pertinent to claim 1 for disclosing a system that analyzes online content to determine difficulty levels and recommend material aligned with a user's skill, which is identified via manual selection or profile analysis of reading history, wherein the system utilizes a learner module to suggest content with a matching or slightly higher difficulty score than previous material, facilitating user improvement, ¶¶ 25 and 107.
Hwang; Jin-young US 20190042551 A1, is pertinent to claim 1 for disclosing AI technology is composed of machine learning (for example, deep learning), ¶ 5, which generates summary information based on user history information, e.g., knowledge level [literacy], ¶¶ 51, 90 and 127.
Sajda; Paul et al. US 20240005398 A1, is pertinent to claim 1 for disclosing a trading platform that generates custom user interfaces for multiple users, ¶ 850, outputting expert trading decisions, Abstract.
Gambhir; Prerana Dharmesh US 20240045581 A1, is pertinent to claim 1 for disclosing a system for optimizing and personalizing the home screen of an application, Abstract, where a user’s profile information may include user preference information, location information, search history, stock tracking history, and the like. Widget types may include weather widgets, social media widgets, productivity application widgets, news widgets, stock widgets, and so forth. Given the known history of a user, such as their browsing history, ¶ 33.
Currell; Nathan et al. US 20250068625 A1, is pertinent to claim 1 for disclosing methods for processing natural language queries, Abstract, including collecting personalized financial information, ¶ 63, and soring chunks of information in a vector database, ¶ 75.
Fleming; Kala et al. US 20190057071 A1, is pertinent to claim 1 for disclosing delivering a summarization of the digital content to the at least one device where the summarizations are tailored [personalized] to a user in an educational setting, Abstract and ¶ 20.
Kumbure, Mahinda Mailagaha et al., Non-Patent Literature, Machine learning techniques and data for stock market forecasting: A literature review (2022), is pertinent to claim 1 for disclosing machine learning techniques that are applied for stock market prediction, Abstract.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to GABRIEL S MERCADO whose telephone number is (408)918-7537. The examiner can normally be reached Mon-Fri 8am-5pm (Eastern Time).
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, Kieu Vu can be reached at (571) 272-4057. 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.
/Gabriel Mercado/Primary Examiner, Art Unit 2171
Prosecution Insights