DETAILED ACTION
1. This communication is in response to the Application filed on 11/10/2023. Claims 1-20 are pending and have been examined.
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.
2. Regarding claims 15-20, the broadest reasonable interpretation of a claim drawn to a computer readable media typically covers forms of non-transitory tangible media and transitory propagating signals per se in view of the ordinary and customary meaning of computer readable storage media. See Subject Matter Eligibility of Computer Readable Media, 1351 OG 212 (26 Jan 2010). See MPEP 2111.01. Signals are nothing but the physical characteristics of a form of energy, and as such is non-statutory natural phenomena. See, e.g., In re Nuitjen, Docket no. 2006-1371 (Fed. Cir. Sept.20, 2007)(slip. op. at 18)("A transitory, propagating signal like Nuitjen's is not a process, machine, manufacture, or composition of matter.' ... Thus, such a signal cannot be patentable subject matter."). Thus, claims 15-20 are rejected under 35 U.S.C. 101 because, giving the claims their broadest reasonable interpretation, the claimed "computer-readable media" encompasses non-statutory subject matter, and is therefore patent-ineligible. It is advised that "computer readable media" be amended to "non-transitory computer readable media" to overcome this rejection, even with SPEC [0076] “In no case is the computer readable storage media a propagated signal.”
3. Claims 1-20 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more. The independent claims 1, 9, 15 recite a method, an apparatus and a computer readable media, thus relating to a statutory category.
Claims 1, 9, 15 recite “observing prompting associated with a large language model service .. identifying insights into the prompting .. enabling display of the insights ..” The limitations as drafted cover mental processes, where a human can observe any prompt such as a question or statement, identify insight of the prompt such as the key points or topic of the question or statement, and display the insight in any manner such as in writing or vocally.
This judicial exception is not integrated into a practical application. In particular, independent claims 1, 9, 15 recite additional elements of “computing device,” “processor” and “computer readable media” which amount to general purpose computing devices. Accordingly, these additional elements do not integrate the abstract idea into a practical application because it does not impose any meaningful limits on practicing the abstract idea. The claims are directed to an abstract idea (SPEC [0075] – Examples of processing system 602 include general purpose central processing units, graphical processing units, application specific processors, and logic devices, as well as any other type of processing device, combinations, or variations thereof).
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 the integration of the abstract idea into a practical application, the additional element of using a processor is noted as a general computer. Mere instructions to apply an exception using a generic computer component cannot provide an inventive concept. Further, the additional limitations in the claims noted above are directed towards insignificant solution activity. The claims are not patent eligible. The applicant is advised to further recite for “a large language model” a particular neural network implementation with specific training to overcome 35 USC 101 abstract idea rejections.
With respect to claims 2, 11, 16, the claims recite “observing replies to the prompts and observing user actions with respect to the replies ..” where a human can observe the LLM service replies/answers to the prompt/question and the resultant user action. The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception.
With respect to claims 3, 12, 17, the claims recite “organizing the prompting into conversations; classifying each of the conversations as belonging to one or more of a set of categories based at least on characteristics of the prompts, characteristics of the replies, and characteristics of the user actions; and identifying trends with respect to the set of categories ..” where a human can perform all of the above recited limitations. The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception.
With respect to claims 4, 18, the claims recite “wherein the categories comprise a subset of categories associated with prompting types .. comprising a creative category, a productivity category, a learning category, and a research category ..” where a human can categorize prompt types into different categories. The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception.
With respect to claims 5, 19, the claims recite “wherein the categories comprise a subset of categories associated with prompting topics .. comprising an off-task category, an on-task category, and an inappropriate content category ..” where a human can categorize prompt topics into different categories. The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception.
With respect to claims 6, 20, the claims recite “wherein the categories comprise a subset of categories associated with prompting quality .. comprising a high-quality category and a low-quality category ..” where a human can categorize prompt quality into different categories. The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception.
With respect to claims 7, 13, the claims recite “wherein the categories comprise: a first subset of categories associated with prompting types, wherein the first subset of categories comprises a creative category, a productivity category, a learning category, and a research category; a second subset of categories associated with prompting topics, wherein the second subset of categories comprises an off-task category, an on-task category, and an inappropriate content category; and a third subset of categories associated with prompting quality, wherein the third subset of categories comprises a high-quality category and a low-quality category ..” where a human can recognize all different categories of prompting types, topics and quality. The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception.
With respect to claims 8, 14, the claims recite “wherein: the characteristics of the prompts comprises content of the prompts; the characteristics of the replies comprises content of the replies; and the characteristics of the user actions comprises dwell time over the replies, a frequency of using a stop-replying feature with respect to the replies, and a frequency of click-throughs with respect to the content in the replies ..” where a human can recognize all different characteristics of the contents of the prompts and the replies, and various characteristics such as related to clicks associated with the user action. The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception.
With respect to claim 10, the claim recites “wherein the insights comprise trends identified in the prompting ..” where a human can observe the prompt/question to identify a trend. The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception.
Claim Rejections - 35 USC § 103
4. Claims 1-2, 9, 11, 15-16 are rejected under 35 U.S.C. 103 as being unpatentable over Bentitou, et al. (US 20180020093; hereinafter BENTITOU) in view of QU, et al. (ASIS&T, 2012.; hereinafter QU) and further in view of Su, et al. (MRQA 2019; hereinafter SU).
As per claim 1, BENTITOU (Title: Automated call answering based on artificial intelligence) discloses “A method of operating [ an insights service ], the method comprising:
observing, by [ an insights service ], prompting associated with [ a large language model service ], wherein the observation is performed on a per-user basis with respect to each user in a group of observed users (BENTITOU, [0157], a caller .. after being prompted with a question <read on ‘prompt’>; [Abstract], A method of processing a telephone call from a calling party <read on ‘a per-user basis .. a group of users> in order to determine the disposition of the call);
identifying, by [ the insights service, insights ] into the prompting on the per-user basis with respect to each of the group of observed users; and enabling, by the insights service, display of the insights in a user interface associated with a reviewing user (BENTITOU, [Abstract] Certain details of an incoming call or of the calling party are obtained by artificial intelligence conversations with the calling party); and
enabling, by the insights service, display of the insights in a user interface associated with a reviewing user (BENTITOU, [Abstract], the artificial intelligence can automatically determine to process the call by .. providing a message or response to the calling party, taking a message from the calling party and appropriately forwarding it to the particular person, to voice mail, or to another person of the business entity <read on ‘a user interface’ and ‘a reviewing user’>; [0164], All the obtained details can be forwarded to and displayed on the called party's transceiver in real-time).”
BENTITOU does not explicitly disclose “an insights service .. insight ..” However, the limitation is taught by QU (Title: An Evaluation of Classification Models for Question Topic Categorization.” Note that [SPEC – 0036] “The insights may include groupings of conversations by topic ..”
In the related field of endeavor, QU teaches: [Introduction, para 2] “Questions in CQA services are organized into hierarchies of categories that often comprise thousands of leaf categories, where each category represents a topic” and [Introduction, para 4] “automatic question topic classification ..”
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate the teachings of QU in the system (as taught by BENTITOU) to identify insights including topics based on input prompts for improved and effective question-answering.
BENTITOU in view of QU does not explicitly disclose “a large language model service ..” However, the limitation is taught by SU (Title: Generalizing Question Answering System with Pre-trained Language Model Fine-tuning).
In the related field of endeavor, SU teaches: [Abstract] “Question answering (QA) systems .. Our model is built on top of a large pre-trained language model, such as XLNet, and then fine-tuned on multiple RC datasets“ where XLNet is an LLM.
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate the teachings of SU in the system (as taught by BENTITOU and QU) to provide an LLM for improved and effective question-answering.
As per claim 2 (dependent on claim 1), BENTITOU in view of QU and SU further discloses “observing, by the insights service, prompts submitted to the large language model service; observing, by the insights service, replies to the prompts from the large language model service; and observing, by the insights service, user actions with respect to the replies (BENTITOU, [0157], a caller .. after being prompted with a question <read on ‘prompt’>; SU, [Abstract], Question answering (QA) systems <where ‘question’ also reads on ‘prompt’> .. Our model is built on top of a large pre-trained language model, such as XLNet; [Abstract], determine the disposition of the call. Certain details of an incoming call or of the calling party are obtained .. providing a message or response to the calling party <read on ‘replies’>; [0018], The recorded audio between the artificial intelligence and the calling party or the conversations themselves be used to generate a transcript which is forwarded to the particular person for present or future action in deciding whether to answer or return the call or not or take other action).”
Claims 9, 15 (both similar in scope to claim 1) are rejected under the same rationale as detailed above for claim 1. Claims 9, 15 further recite computer readable storage media and processor (BENTITOU, [0044], The computer system is a combination of a tangible storage device, processor, and other hardware components that carry out instructions to imitate the manners in which a human receives audio or text, processes the received audio or text, and provides a response related to the audio or text).
Claims 11, 16 (similar in scope to claim 2) are rejected under the same rationale as detailed above for claim 2.
5. Claims 3, 10, 12, 17 are rejected under 35 U.S.C. 103 as being unpatentable over BENTITOU in view of QU and SU, and further in view of Benhardus (2010 UCCS REU FOR ARTIFICIAL INTELLIGENCE; hereinafter BENHARDUS).
As per claim 3 (dependent on claim 2), BENTITOU in view of QU and SU, further discloses “organizing, by the insights service, the prompting into conversations; classifying, by the insights service, each of the conversations as belonging to one or more of a set of categories based at least on characteristics of the prompts, characteristics of the replies, and characteristics of the user actions; and [ identifying, by the insights service, trends ] with respect to the set of categories (BENTITOU, [0018], The recorded audio between the artificial intelligence and the calling party or the conversations <read on ‘organizing the prompting into conversations’>; QU, [Introduction, para 2], Questions in CQA services are organized into hierarchies of categories that often comprise thousands of leaf categories; [Introduction, para 4], automatic question topic classification <read on ‘classifying .. each of the conversations .. based at least on characteristics of the prompts, characteristics of the replies, and characteristics of the user actions’ where topic is classified based on all question/response relevant information>).”
BENTITOU in view of QU and SU does not explicitly disclose “identifying, by the insights service, (conversation) trends ..” However, the limitation is taught by BENHARDUS (Title: Streaming Trend Detection in Twitter).”
In the related field of endeavor, BENHARDUS teaches: [Abstract] “detecting and identifying trending topics from streaming data.”
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate the teachings of BENHARDUS in the system (as taught by BENTITOU, QU and SU) to classify conversations into different categories based on conversation trends for improved and effective question-answering.
As per claim 10 (dependent on claim 9), BENTITOU in view of QU and SU further discloses “wherein the insights comprise [ trends identified in the prompting ] based on observations by the insights service of the prompting).”
BENTITOU in view of QU and SU does not explicitly disclose “trends identified in the prompting ..” However, the limitation is taught by BENHARDUS (Title: Streaming Trend Detection in Twitter).”
In the related field of endeavor, BENHARDUS teaches: [Abstract] “detecting and identifying trending topics from streaming data” where streaming data” reads on “input data” or “prompts.”
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate the teachings of BENHARDUS in the system (as taught by BENTITOU, QU and SU) to classify conversations into different categories based on conversation trends for improved and effective question-answering.
Claim 12 (similar in scope to claim 3) is rejected under the same rationale as detailed above for claim 3.
Claim 17 (similar in scope to claim 3) is rejected under the same rationale as detailed above for claim 3.
5. Claims 4, 18 rejected under 35 U.S.C. 103 as being unpatentable over BENTITOU in view of QU, SU and BENHARDUS, and further in view of Sangodiah, et al. (JATIT, 2015; hereinafter SANGODIAH).
As per claim 4 (dependent on claim 3), BENTITOU in view of QU, SU and BENHARDUS further discloses “wherein the categories comprise a subset of categories associated with [prompting types ], the subset of categories comprising a creative category, a productivity category, a learning category, and a research category.
BENTITOU in view of QU, SU and BENHARDUS does not explicitly disclose “prompting types ..” However, the limitation is taught by SANGODIAH (Title: QUESTION CLASSIFICATION USING STATISTICAL APPROACH: A COMPLETE REVIEW).
In the related field of endeavor, SANGODIAH teaches: [Introduction, para 7] “question classification in QA and IR .. factual type of questions and not complex questions .. question classification in educational environment .. imperative and factual types of questions .. question classification involving other languages ..” which teaches numerous various question/prompting types and the associated applications, including creative, productivity, learning and research types and applications.
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate the teachings of SANGODIAH in the system (as taught by BENTITOU, QU, SU and BENHARDUS) to classify conversations into different categories based on prompt types for improved and effective question-answering.
Claim 18 (similar in scope to claim 4) is rejected under the same rationale as detailed above for claim 4.
6. Claims 5, 7, 13, 19 are rejected under 35 U.S.C. 103 as being unpatentable over BENTITOU in view of QU, SU and BENHARDUS, and further in view of Vasilyeva (EJPC, 2012; hereinafter VASILYEVA).
As per claim 5 (dependent on claim 3), BENTITOU in view of QU, SU and BENHARDUS further discloses “wherein the categories comprise a subset of categories associated with prompting topics, the subset of categories comprising [ an off-task category, an on-task category, and an inappropriate content category ] (QU, [Introduction, para 2], Questions in CQA services are organized into hierarchies of categories that often comprise thousands of leaf categories <where thousands of leaf categories read on any of huge number of different leaf categories>, where each category represents a topic; [Introduction, para 4], automatic question topic classification).”
BENTITOU in view of QU, SU and BENHARDUS does not explicitly disclose “an off-task category, an on-task category, and an inappropriate content category ..” However, the limitation is taught by VASILYEVA (Title: Topics as indication of being on-task/off-task in dispute mediation).
In the related field of endeavor, VASILYEVA teaches: [Abstract] “examines topics that participants discuss in the course of mediation sessions in order to understand how these topics indicate whether a mediation session is on-task or off-task .. by introducing institutionally appropriate topics and terminating off-task ones <read on including ‘inappropriate content’>.”
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate the teachings of VASILYEVA in the system (as taught by BENTITOU, QU, SU and BENHARDUS) to classify conversations into different categories based on prompt topics for improved and effective question-answering.
As per claim 7 (dependent on claim 3), BENTITOU in view of QU, SU and BENHARDUS further discloses “wherein the categories comprise: a first subset of categories associated with prompting types, wherein the first subset of categories comprises a creative category, a productivity category, a learning category, and a research category; [ a second subset of categories associated with prompting topics, wherein the second subset of categories comprises an off-task category, an on-task category, and an inappropriate content category ]; and a third subset of categories associated with prompting quality, wherein the third subset of categories comprises a high-quality category and a low-quality category.”
BENTITOU in view of QU, SU and BENHARDUS does not explicitly disclose “a second subset of categories associated with prompting topics, wherein the second subset of categories comprises an off-task category, an on-task category, and an inappropriate content category ..” However, the limitation is taught by VASILYEVA (Title: Topics as indication of being on-task/off-task in dispute mediation).
In the related field of endeavor, VASILYEVA teaches: [Abstract] “examines topics that participants discuss in the course of mediation sessions in order to understand how these topics indicate whether a mediation session is on-task or off-task .. by introducing institutionally appropriate topics and terminating off-task ones.”
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate the teachings of VASILYEVA in the system (as taught by BENTITOU, QU, SU and BENHARDUS) to classify conversations into different categories based on prompt topics for improved and effective question-answering.
Claims 13, 19 (similar in scope to claims 7, 5, respectively) are rejected under the same rationale as detailed above for claims 7, 5, respectively.
7. Claims 6, 20 are rejected under 35 U.S.C. 103 as being unpatentable over BENTITOU in view of QU, SU and BENHARDUS, and further in view of Shah, et al. (HICSS, 2014; hereinafter SHAH).
As per claim 6 (dependent on claim 3), BENTITOU in view of QU, SU and BENHARDUS further discloses “wherein the categories comprise a subset of categories associated with [prompting quality ], the subset of categories comprising a high-quality category and a low-quality category.
BENTITOU in view of QU, SU and BENHARDUS does not explicitly disclose “prompting quality ..” However, the limitation is taught by SHAH (Title: Questioning the Question – Addressing the Answerability of Questions in Community Question-Answering).
In the related field of endeavor, SHAH teaches: [Abstract] “we investigate question quality among questions posted in Yahoo! Answers to assess what factors contribute to the goodness of a question and determine if we can flag poor quality questions. Using human assessments of whether a question is good or bad and extracted textual features from the questions,”
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate the teachings of SHAH in the system (as taught by BENTITOU, QU, SU and BENHARDUS) to classify conversations into different categories based on prompt quality for improved and effective question-answering.
Claim 20 (similar in scope to claim 6) is rejected under the same rationale as detailed above for claim 6.
8. Claims 8, 14 are rejected under 35 U.S.C. 103 as being unpatentable over BENTITOU in view of QU, SU and BENHARDUS, and further in view of Kim, et al. (WSDM’14, 2014; hereinafter KIM).
As per claim 8 (dependent on claim 3), BENTITOU in view of QU, SU and BENHARDUS further discloses “wherein: the characteristics of the prompts comprises content of the prompts; the characteristics of the replies comprises content of the replies (BENTITOU, [Abstract], processing a telephone call .. to determine the disposition of the call. Certain details of an incoming call .. are obtained <read on characteristics and contents of ‘prompt’> .. providing a message or response to the calling party <read on characteristics and contents of ‘replies’>); and [ the characteristics of the user actions comprises dwell time over the replies, a frequency of using a stop-replying feature with respect to the replies, and a frequency of click-throughs with respect to the content in the replies ].
BENTITOU in view of QU, SU and BENHARDUS does not explicitly disclose “the characteristics of the user actions comprises dwell time over the replies, a frequency of using a stop-replying feature with respect to the replies, and a frequency of click-throughs with respect to the content in the replies ..” However, the limitation is taught by KIM (Title: Modeling Dwell Time to Predict Click-level Satisfaction).
In the related field of endeavor, KIM teaches: [Abstract] “Clicks <read on ‘click-throughs’ or ‘stop-replying’ – no clicking, and all associated information such as frequency for user reactions to system replies> on search results <read on ‘replies’> are the most widely used behavioral signals <read on ‘user actions’> for predicting search satisfaction .. A popular heuristic .. is to only consider clicks with long dwell time, usually equaling or exceeding 30 seconds. The rationale is that the more time a searcher spends on a page, the more likely they are to be satisfied with its contents.”
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate the teachings of KIM in the system (as taught by BENTITOU, QU, SU and BENHARDUS) to provide user actions/reactions to the system replies including click-throughs, non-clicking (non-replies) and different dwell time (long or short) over the replies for responsive, interactive question-answering.
Claim 14 (similar in scope to claim 8) is rejected under the same rationale as detailed above for claim 8.
Conclusion
9. Any inquiry concerning this communication or earlier communications from the examiner should be directed to FENG-TZER TZENG whose telephone number is 571-272-4609. The examiner can normally be reached on M-F (9:00-5:00). The fax phone number where this application or proceeding is assigned is 571-273-4609.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Paras Shah (SPE) can be reached on 571-270-1650.
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.
/FENG-TZER TZENG/ 2/20/2026
Primary Examiner, Art Unit 2653