DETAILED ACTION
This Office Action is in response to the correspondence filed by the applicant on 12/30/2024.
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 .
Information Disclosure Statement
The Information Statements (IDS) filed on 1/3/2025 have been accepted and considered in this office action and are in compliance with the provisions of 37 CFR 1.97.
Allowable Subject Matter
Claims 2, 9, and 16 are objected to as being dependent upon a rejected base claim, but would be allowable if rewritten in independent form including all of the limitations of the base claim and any intervening claims.
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.
Claims 1, 4-6, 8, 11-13, 15, and 18-20 are rejected under 35 U.S.C. 103 as being unpatentable over DESAI (US 2024/0152704 A1), and in further view of DEMPSEY (US 12,113,934 B1).
REGARDING CLAIM 1, DESAI discloses a computer-implemented method comprising:
receiving, by one or more processors, streaming data indicating a set of digital interactions between a first user and a second user (DESAI Par 86 – “Thereafter, the collaboration platform can acquire, from the multiple computing devices, (i) streams of voice data that include words spoken by the multiple individuals as part of the conversation and (ii) streams of interactivity data that indicate spatial interactions of the multiple individuals with the virtual environment (step 702). Thus, the collaboration platform can receive a separate stream of voice data and a separate stream of interactivity data from each of the multiple computing devices. For each of the multiple participants, the corresponding stream of interactivity data can indicate the actual or desired spatial interactions of an avatar corresponding to that user with avatars associated with other participants in the conversation.”);
predicting, by an intent shift classification model and based at least in part on the streaming data, a span of time (DESAI Par 63 – “The phrase recognizer 412 may be responsible for determining, based on a comparison of the transcript 414 to a pre-supplied library 422 of phrases that indicate a beginning, ending, or changing of topic.”; Par 69 – “Referring to the former, the phrase recognizer 412 can employ artificial intelligence and semantic analysis to discover a phrase that indicates a participant 406 wishes to delineate the end of the current knowledge object from the beginning of the next knowledge object. For example, the participant can utter “Let's move on to the next topic,” and when this phrase—or a semantically identical or comparable phrase—is detected, the phrase recognizer 412 ma can y “cut” the rolling scroll 424 so that the current knowledge object is finalized and its card is added to the stack 428.”) over which a portion of the set of digital interactions are associated with a first intent (DESAI Par 88 – “The collaboration platform can then segment the transcript based on a semantic analysis of (i) words contained therein and (ii) the streams of interactivity data revealing that multiple topics are covered in the conversation (step 704). For example, the collaboration platform can segment the textual stream into multiple segments by statistically evaluating the textual stream using learnt statistical models associated with a set of phrases, so as to classify text therein using the learnt statistical models. As another example, the collaboration platform can segment the textual stream into multiple segments based on a comparison of its words to (i) an agenda provided for the conversation, (ii) a template identified for the conversation, or (iii) a set of phrases indicative of a desire to delineate between different topics. Assume, for example, that the collaboration platform determines that the conversation has shifted from a first topic to a second topic by comparing the transcript to a template identified for the conversation. In response to such a determination, the collaboration platform can conclude the current transcript segment that corresponds to the first topic and commence the next transcript segment that corresponds to the second topic. Each of the multiple topics can be associated with a corresponding one of multiple transcript segments.”);
classifying, by an intent classification model and based at least in part on the portion of the set of digital interactions, a first intent classification associated with the portion of the set of digital interactions (DESAI Par 18 –“ Discourse throughout a conversation can be converted into a transcription (or simply “transcript”), parsed to identify topical shifts, and then segmented based on the topical shifts. Together, the various topics discussed over the course of a conversation may be called a “topic listing,” “topic cloud,” or “mind map.””; Par 62 – “Semantic analysis could also be performed by the transcription service—or some other service—and therefore, the collaboration platform may not include the semantic analyzer 410 in some embodiments. In embodiments where the collaboration platform does not include a semantic analyzer 410, the collaboration platform may obtain (e.g., from the transcription service or another service) the topic sets 416, insights, and summaries 420.”; Par 93 – “As an example, topics can be identified based on a semantic analysis of the transcript. These topics may be representative of insights into the conversation, and therefore can be posted in the virtual environment for review by the participants, such that the participants can readily determine which topics have been discussed.”);
[generating, by the one or more processors, a first prompt based at least in part on (i) the first intent classification and (ii) the portion of the set of digital interactions];
generating, [by a generative machine-learned model and based at least in part on the prompt], a first summary of the set of digital interactions (DESAI Pars 64-67 – “As further discussed below, the collaboration platform can record content of the conversation and contextual metadata related to the conversation into the current knowledge object. For example, the current knowledge object can include: A list of participants who are engaged in the conversation or who have contributed to the corresponding topic; Outputs from processing of the voice data 404, such as the transcript 414, topic sets 416, insights such as questions and action items 418, and summary 420; and”; Par 91 – “Further, summaries of the topics can be posted for review. In FIG. 8B, a summary has been added beneath the first topic. This summary can be extracted, inferred, or otherwise derived from the transcript (and more specifically, the segment of the transcript that corresponds to the first topic). FIG. 8C illustrates how a summary can be added beneath the second topic to account for contribution by a second speaker, while FIG. 8D illustrates how a summary can be added beneath the third topic to account for contribution by a third speaker. As discussed above, these summaries can be output by a semantic analyzer (e.g., semantic analyzer 410 of FIG. 4 ).”); and
causing, by the one or more processors, the first summary of the set of digital interactions to be displayed in association with the portion of the set of digital interactions (DESAI Par 91 – “Further, summaries of the topics can be posted for review. In FIG. 8B, a summary has been added beneath the first topic. This summary can be extracted, inferred, or otherwise derived from the transcript (and more specifically, the segment of the transcript that corresponds to the first topic). FIG. 8C illustrates how a summary can be added beneath the second topic to account for contribution by a second speaker, while FIG. 8D illustrates how a summary can be added beneath the third topic to account for contribution by a third speaker. As discussed above, these summaries can be output by a semantic analyzer (e.g., semantic analyzer 410 of FIG. 4 ).”).
DESAI does not explicitly teach the [square-bracketed] limitations. In other words, DESAI teaches generating a summary for interactive conversation, but DESAI does not explicitly teach using an LLM and its input, a prompt.
DEMPSEY discloses the [square-bracketed] limitations. DEMPSEY discloses a method/system for analyzing intelligent call for service interaction comprising:
[generating, by the one or more processors, a first prompt based at least in part on (i) the first intent classification and (ii) the portion of the set of digital interactions] (DEMPSEY Col 7:20-62 – “In some embodiments, the postprocessing module 122 can utilize the LLM (or another machine learning model) to generate a summary of the call transcript. For example, the postprocessing module 122 can provide as inputs to the LLM (i) the call transcript and (ii) a specific request/prompt to generate a summary for the call transcript. … . In some embodiments, the summary generated by the LLM is a topic-based summary. To generate such a summary, the postprocessing module 122 first asks the LLM to identify the call topic by supplying as inputs to the LLM the call transcript and the prompt “What is the topic of this call?” In response, the LLM returns the call topic identified from the transcript (e.g., “Hardship Withdrawal”). Alternatively, the call topic can be the same as the topic identified at step 206 of process 200. The postprocessing module 122 can then customize the prompt/question to the LLM by asking the LLM to summarize the call transcript with respect to the business requirements for the call topic identified (e.g., “summarize call and include why a hardship withdrawal is being taken”). In some embodiments, the summary generated by the LLM is a user-based summary, which is typically employed for a frequent customer or call agent to determine a conversation pattern.”);
generating, [by a generative machine-learned model and based at least in part on the prompt], a first summary of the set of digital interactions (DEMPSEY Col 7:20-62 – “In some embodiments, the postprocessing module 122 can utilize the LLM (or another machine learning model) to generate a summary of the call transcript. For example, the postprocessing module 122 can provide as inputs to the LLM (i) the call transcript and (ii) a specific request/prompt to generate a summary for the call transcript. In return, the LLM provides a summary of the call transcript as requested using, for example, the zero-shot technique described above.”); and
causing, by the one or more processors, the first summary of the set of digital interactions to be displayed in association with the portion of the set of digital interactions (DEMPSEY Col 7:63-8:6 – “In some embodiments, the postprocessing module 122 can also cause to display on the graphical user interface the summary of the call transcript generated by the LLM. Exemplary graphical user interfaces are described below with reference to FIGS. 3-6 .”).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the method/system of DESAI to include an LLM and prompt generation for the LLM, as taught by DEMPSEY.
One of ordinary skill would have been motivated to include an LLM and prompt generation for the LLM, in order to provide a more accurate summarization using the state of the art technology in the field of natural language processing.
REGARDING CLAIM 4, DESAI in view of DEMPSEY discloses the computer-implemented method of claim 1, wherein predicting the span of time (DESAI Par 63 – “The phrase recognizer 412 may be responsible for determining, based on a comparison of the transcript 414 to a pre-supplied library 422 of phrases that indicate a beginning, ending, or changing of topic.”; Par 69 – “Referring to the former, the phrase recognizer 412 can employ artificial intelligence and semantic analysis to discover a phrase that indicates a participant 406 wishes to delineate the end of the current knowledge object from the beginning of the next knowledge object. For example, the participant can utter “Let's move on to the next topic,” and when this phrase—or a semantically identical or comparable phrase—is detected, the phrase recognizer 412 ma can y “cut” the rolling scroll 424 so that the current knowledge object is finalized and its card is added to the stack 428.”) over which a portion of the set of digital interactions are associated with a first intent includes predicting the span of time over which a portion of the set of digital interactions (DESAI Par 88 – “The collaboration platform can then segment the transcript based on a semantic analysis of (i) words contained therein and (ii) the streams of interactivity data revealing that multiple topics are covered in the conversation (step 704). For example, the collaboration platform can segment the textual stream into multiple segments by statistically evaluating the textual stream using learnt statistical models associated with a set of phrases, so as to classify text therein using the learnt statistical models. As another example, the collaboration platform can segment the textual stream into multiple segments based on a comparison of its words to (i) an agenda provided for the conversation, (ii) a template identified for the conversation, or (iii) a set of phrases indicative of a desire to delineate between different topics. Assume, for example, that the collaboration platform determines that the conversation has shifted from a first topic to a second topic by comparing the transcript to a template identified for the conversation. In response to such a determination, the collaboration platform can conclude the current transcript segment that corresponds to the first topic and commence the next transcript segment that corresponds to the second topic. Each of the multiple topics can be associated with a corresponding one of multiple transcript segments.”) are associated with a first intent as a start of a new intent or an end of a current intent (DESAI Par 69 – “Referring to the former, the phrase recognizer 412 can employ artificial intelligence and semantic analysis to discover a phrase that indicates a participant 406 wishes to delineate the end of the current knowledge object from the beginning of the next knowledge object. For example, the participant can utter “Let's move on to the next topic,” and when this phrase—or a semantically identical or comparable phrase—is detected, the phrase recognizer 412 ma can y “cut” the rolling scroll 424 so that the current knowledge object is finalized and its card is added to the stack 428.”).
REGARDING CLAIM 5, DESAI in view of DEMPSEY discloses the computer-implemented method of claim 4, wherein at least one of:
classifying the first portion of the set of digital interactions as the first intent shift (DESAI Par 88 – “For example, the collaboration platform can segment the textual stream into multiple segments by statistically evaluating the textual stream using learnt statistical models associated with a set of phrases, so as to classify text therein using the learnt statistical models. As another example, the collaboration platform can segment the textual stream into multiple segments based on a comparison of its words to (i) an agenda provided for the conversation, (ii) a template identified for the conversation, or (iii) a set of phrases indicative of a desire to delineate between different topics. Assume, for example, that the collaboration platform determines that the conversation has shifted from a first topic to a second topic by comparing the transcript to a template identified for the conversation. In response to such a determination, the collaboration platform can conclude the current transcript segment that corresponds to the first topic and commence the next transcript segment that corresponds to the second topic. Each of the multiple topics can be associated with a corresponding one of multiple transcript segments.”) includes classifying the first portion of the interaction as the start of the new intent, the first span begins at a first digital interaction set of digital interactions; or classifying the first portion of the set of digital interactions as the first intent shift includes classifying the first portion of the set of digital interactions as the end of the current intent, the first span ends at the first portion of the set of digital interactions (DESAI Par 63 – “The phrase recognizer 412 may be responsible for determining, based on a comparison of the transcript 414 to a pre-supplied library 422 of phrases that indicate a beginning, ending, or changing of topic.”; Par 69 – “Referring to the former, the phrase recognizer 412 can employ artificial intelligence and semantic analysis to discover a phrase that indicates a participant 406 wishes to delineate the end of the current knowledge object from the beginning of the next knowledge object. For example, the participant can utter “Let's move on to the next topic,” and when this phrase—or a semantically identical or comparable phrase—is detected, the phrase recognizer 412 ma can y “cut” the rolling scroll 424 so that the current knowledge object is finalized and its card is added to the stack 428.”).
REGARDING CLAIM 6, DESAI in view of DEMPSEY discloses the computer-implemented method of claim 1, further comprising:
receiving, by the one or more processors, [service] interaction data indicating the second user's interaction with a hosted service (DESAI Par 86 – “Thereafter, the collaboration platform can acquire, from the multiple computing devices, (i) streams of voice data that include words spoken by the multiple individuals as part of the conversation and (ii) streams of interactivity data that indicate spatial interactions of the multiple individuals with the virtual environment (step 702). Thus, the collaboration platform can receive a separate stream of voice data and a separate stream of interactivity data from each of the multiple computing devices. For each of the multiple participants, the corresponding stream of interactivity data can indicate the actual or desired spatial interactions of an avatar corresponding to that user with avatars associated with other participants in the conversation.”); and
wherein predicting the span of time (DESAI Par 63 – “The phrase recognizer 412 may be responsible for determining, based on a comparison of the transcript 414 to a pre-supplied library 422 of phrases that indicate a beginning, ending, or changing of topic.”; Par 69 – “Referring to the former, the phrase recognizer 412 can employ artificial intelligence and semantic analysis to discover a phrase that indicates a participant 406 wishes to delineate the end of the current knowledge object from the beginning of the next knowledge object. For example, the participant can utter “Let's move on to the next topic,” and when this phrase—or a semantically identical or comparable phrase—is detected, the phrase recognizer 412 ma can y “cut” the rolling scroll 424 so that the current knowledge object is finalized and its card is added to the stack 428.”) over which a portion of the set of digital interactions is associated with a first intent is based at least in part on the [service] interaction data (DESAI Par 88 – “The collaboration platform can then segment the transcript based on a semantic analysis of (i) words contained therein and (ii) the streams of interactivity data revealing that multiple topics are covered in the conversation (step 704). For example, the collaboration platform can segment the textual stream into multiple segments by statistically evaluating the textual stream using learnt statistical models associated with a set of phrases, so as to classify text therein using the learnt statistical models. As another example, the collaboration platform can segment the textual stream into multiple segments based on a comparison of its words to (i) an agenda provided for the conversation, (ii) a template identified for the conversation, or (iii) a set of phrases indicative of a desire to delineate between different topics. Assume, for example, that the collaboration platform determines that the conversation has shifted from a first topic to a second topic by comparing the transcript to a template identified for the conversation. In response to such a determination, the collaboration platform can conclude the current transcript segment that corresponds to the first topic and commence the next transcript segment that corresponds to the second topic. Each of the multiple topics can be associated with a corresponding one of multiple transcript segments.”), wherein classifying the intent classification associated with the portion of the set of digital interactions is based at least in part on the [service] interaction data (DESAI Par 18 –“ Discourse throughout a conversation can be converted into a transcription (or simply “transcript”), parsed to identify topical shifts, and then segmented based on the topical shifts. Together, the various topics discussed over the course of a conversation may be called a “topic listing,” “topic cloud,” or “mind map.””; Par 62 – “Semantic analysis could also be performed by the transcription service—or some other service—and therefore, the collaboration platform may not include the semantic analyzer 410 in some embodiments. In embodiments where the collaboration platform does not include a semantic analyzer 410, the collaboration platform may obtain (e.g., from the transcription service or another service) the topic sets 416, insights, and summaries 420.”; Par 93 – “As an example, topics can be identified based on a semantic analysis of the transcript. These topics may be representative of insights into the conversation, and therefore can be posted in the virtual environment for review by the participants, such that the participants can readily determine which topics have been discussed.”).
DESAI does not explicitly teach the [square-bracketed] limitation. In other words, DESAI teaches using interaction data of multiple users, but does not explicitly teach the interaction data is [service] interaction data.
DEMPSEY discloses the [square-bracketed] limitations. DEMPSEY discloses a method/system for analyzing interaction data, wherein the interaction data includes [service] interaction data (DEMPSEY Col 5:14-20 – “The process 200 starts at step 202 with the evaluation system 100 receiving an audio recording of a call between a customer and a call agent.”).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the method/system of DESAI to include service interaction data, as taught by DEMPSEY.
One of ordinary skill would have been motivated to include service interaction data, in order to unlock new insights into customer-agent interactions on quality and compliance standard (DEMPSEY Col 1).
REGARDING CLAIM 8, DESAI in view of DEMPSEY discloses a system comprising: one or more processors; and one or more memories storing processor-executable instructions that, when executed by the one or more processors, cause the one or more processors to perform operations (DESAI Fig. 2; Par 32 – “a processor”) comprising: performing the steps of claim 1; thus, it is rejected under the same rationale.
Claim 11 is similar to claim 4; thus, it is rejected under the same rationale.
Claim 12 is similar to claim 5; thus, it is rejected under the same rationale.
Claim 13 is similar to claim 6; thus, it is rejected under the same rationale.
REGARDING CLAIM 15, DESAI in view of DEMPSEY discloses one or more non-transitory computer-readable media storing processor- executable instructions that, when executed by one or more processors, cause the one or more processors to perform operations comprising: performing the steps of claim 1; thus, it is rejected under the same rationale.
Claim 18 is similar to claim 4; thus, it is rejected under the same rationale.
Claim 19 is similar to claim 5; thus, it is rejected under the same rationale.
Claim 20 is similar to claim 6; thus, it is rejected under the same rationale.
Claims 3, 7, 10, 14, and 17 are rejected under 35 U.S.C. 103 as being unpatentable over DESAI (US 2024/0152704 A1) in view of DEMPSEY (US 12,113,934 B1), and in further view of CHOPRA (US 2023/0112369 A1).
REGARDING CLAIM 3, DESAI in view of DEMPSEY discloses the computer-implemented method of claim 1.
DEMPSEY further teaches a method/system for analyzing service interaction data for generating a summary using an LLM and prompts comprising: receiving, by the one or more processors, a [user] revision of the first intent classification [via a user interface] (DEMPSEY Fig. 2 – “revise one or more inputs to the LLM to improve accuracy 218.”; Col 8:59-9:3 – “In another optional step 218 of process 200, the feedback module 124 of the call agent evaluation system 100 is configured to collect ratings on certain outputs produced by the LLM, including, but are not limited to, (i) the scores to the questions under one or more criteria for evaluating the call agent generated by the LLM at step 210, (ii) the summaries of call transcripts generated by the LLM at step 212, and/or (iii) the performance coaching recommendations generated by the LLM at step 216. These ratings represent a ground truth database that can be used by the feedback module 124 to fine tune, in a feedback loop, various inputs to the LLM to improve evaluation accuracy.”); and generating, by the one or more processors, a second prompt based at least in part on (i) the [user] revised first intent classification and (ii) the portion of the set of digital interactions; (DEMPSEY Fig. 2 – “revise one or more inputs to the LLM to improve accuracy 218.”; Col 7:20-62 – “In some embodiments, the postprocessing module 122 can utilize the LLM (or another machine learning model) to generate a summary of the call transcript. For example, the postprocessing module 122 can provide as inputs to the LLM (i) the call transcript and (ii) a specific request/prompt to generate a summary for the call transcript. … . In some embodiments, the summary generated by the LLM is a topic-based summary.”).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the method/system of DESAI to include an LLM and prompt generation for the LLM, as taught by DEMPSEY.
One of ordinary skill would have been motivated to include an LLM and prompt generation for the LLM, in order to provide a more accurate summarization using the state of the art technology in the field of natural language processing.
However, DESAI in view of DEMPSEY does not explicitly teach the [square-bracketed] limitation. In other words, DEMPSEY teaches revising a prompt to an LLM to improve its performance, but does not explicitly teach the revision is based on an input from a user.
CHOPRA discloses the [square-bracketed] limitations. CHOPRA discloses a method/system for analyzing service interaction data for classifying intent and generating a summary further comprising: receiving, by the one or more processors, a user revision of the first intent classification via a user interface (CHOPRA Fig. 5; Par 62 – “The client computing device 103 can receive feedback (step 404) from the agent/CSR by enabling the CSR to modify the text contained in the intent and/or short summary fields 506, 508 in order to correct and/or enhance the data contained therein. For example, the model 108 a could have generated an intent for the call that reads ‘updated address,’ e.g. via misinterpretation of the computer text segment. Because the actual customer intent for the call was to update his phone number, the CSR can edit the text in the intent field 506 to reflect the correct intent.“);
generating, by the one or more processors, a second prompt based at least in part on (i) the [user] revised first intent classification (CHOPRA Fig. 5; Par 62 – “The client computing device 103 can receive feedback (step 404) from the agent/CSR by enabling the CSR to modify the text contained in the intent and/or short summary fields 506, 508 in order to correct and/or enhance the data contained therein.”) and (ii) the portion of the set of digital interactions (CHOPRA Figs. 3 and 4 shows feeding the agent feedback to the system for generating the summary and intent and DEMPSEY already teaches generating a prompt based on the revised input and feeding it to an LLM. Thus, the combination teaches generating a second prompt (e.g., a revised input to the system based on the feedback and the interaction).);
generating, by the generative machine-learned model and based at least in part on the second prompt, an updated summary of the set of digital interactions (CHOPRA Figs. 3 and 4; Par 62 – “In some embodiments, after the CSR has saved the changes to output database 102 b, the updated intent data and short summary data is transmitted back to the embedding generation module 110 so that the updated data can be used to train the respective models 108 a, 108 b in order to improve the accuracy of the models (see FIG. 3 ). In this manner, the system 100 advantageously leverages CSR feedback to refine the intent generation and short summary generation processes to result in more accurate and complete intents and short summaries.”;); and
causing, by the one or more processors, the updated summary of the set of digital interactions to be displayed in association with the portions of the set of digital interactions (CHOPRA Figs. 3 and 4; Par 60 – “The client computing device 103 generates a graphical user interface that displays (step 402) the generated intent and/or short summary as created by the models 108 a, 108 b from the computer text segment that contains the interaction data.”).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the method/system of DESAI in view of DEMPSEY to include a user revised intent for generating an updated summary, as taught by DEMPSEY.
One of ordinary skill would have been motivated to include a user revised intent for generating an updated summary, in order to refine the system to result in more accurate and complete intents and summaries.
REGARDING CLAIM 7, DESAI in view of DEMPSEY discloses the computer-implemented method of claim 1.
DESAI further teaches training data for a intent shift classification model (DESAI Par 79 – “As another example, the collaboration platform can segment the textual stream into multiple segments. This can be accomplished in several ways. For example, the collaboration platform can segment the textual stream into multiple segments by statistically evaluating the textual stream using learnt statistical models associated with a set of phrases, so as to classify text therein using the learnt statistical models. Each statistical model can be learned through analysis of training data that includes examples of the corresponding phrase. From the training data, each statistical model is able to learn when use of the corresponding phrase is indicative of a topical shift.”), but does not explicitly teach receiving and generating the training data.
CHOPRA discloses a method/system for classifying an intent and generating a summary further comprising:
receiving, by the one or more processors, service interaction data indicating the second user's interaction with a hosted service (CHOPRA Fig. 2; Par 30 – “The user activity database 102 a includes historical user activity data, which in some embodiments is a dedicated section of the database 102 a that contains specialized data used by the other components of the system 100 to perform at least a portion of the process of automated analysis of customer interaction text to generate customer intent information and a hierarchy of customer issues as described herein. Generally, the historical user activity data comprises data elements associated with prior interactions between one or more end users (e.g., customers) and one or more agents (such as customer service representatives (CSRs)).”); and
generating, by the one or more processors, one or both of (i) training data for the intent shift classification model, and (ii) training data for the intent classification model, based on the service interaction data (CHOPRA Fig. 2 Intent Model 108a Train Model 310; Par 57 – “As shown in FIG. 3 , the intent model 108 a uses at least a portion of the word embeddings received from the module 110 as training data—in that the model 108 a uses embeddings created from existing interaction transcript data and CSR notes data to train (310 a) the model to generate intents that are then confirmed as accurate with corresponding existing intent data for the interactions within an accuracy tolerance.”; Par 61 – “The intent area 506 and the short summary area 508 contain the intent and short summary, respectively, that were generated from the computer text segment 504 by the models 108 a, 108 b of server computing device 106. As shown in FIG. 5 , the intent 506 generated by the model 108 a is ‘updated phone number’ and the short summary generated by the model 108 b is ‘updated phone number and enrolled in voice capture.’”).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the method/system of DESAI in view of DEMPSEY to include generating training data based on service interaction data, as taught by CHOPRA.
One of ordinary skill would have been motivated to include generating training data based on service interaction data, in order to obtain a model tailored for a specific field as desired.
Claim 10 is similar to claim 3; thus, it is rejected under the same rationale.
Claim 14 is similar to claim 7; thus, it is rejected under the same rationale.
Claim 17 is similar to claim 3; thus, it is rejected under the same rationale.
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to JONATHAN C KIM whose telephone number is (571)272-3327. The examiner can normally be reached Monday to Friday 8:00 AM thru 4:00 PM EST.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Andrew C Flanders can be reached at 571-272-7516. 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.
/JONATHAN C KIM/Primary Examiner, Art Unit 2655
Prosecution Insights