DETAILED ACTION
Continued Examination Under 37 CFR 1.114
A request for continued examination under 37 CFR 1.114, including the fee set forth in 37 CFR 1.17(e), was filed in this application after final rejection. Since this application is eligible for continued examination under 37 CFR 1.114, and the fee set forth in 37 CFR 1.17(e) has been timely paid, the finality of the previous Office action has been withdrawn pursuant to 37 CFR 1.114. Applicant's submission filed on 10/22/2025 has been entered. Claims 1, 4-8, 11-15 and 18-26 are pending in the application and have been examined
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Response to Amendment
The response filed on 10/22/2025 has been correspondingly accepted and considered in this Office Action. Claims 1, 4-8, 11-15 and 18-26 have been examined.
Response to Arguments
Applicant's arguments filed 10/22/2025 have been fully considered as follows:
Applicant’s arguments with respect to claim 1 (also representative of claims 8 and 15) state that
“The rejection admits that Nowak and Mody fails to teach that discourse feature
includes at least one or more interruptions in the speech segment. The rejection relies on
Shires… Examiner's assertion that determining importance of user's speech to other users constituting determining a drift in topic in the claimed fashion is not only inconsistent with the teaching of Shires but appears to an assertion based on "broadest possible interpretation" as opposed to "broadest reasonable interpretation.". ”
Applicant’s arguments with respect to claim(s) 1, 8 and 15 have been considered but are moot because the new ground of rejection does not rely on any reference applied in the prior rejection of record for any teaching or matter specifically challenged in the argument.
Applicant’s arguments with respect to claim 22, 24 state that
“The rejection relies on Deng to show the above recited features. Deng, however,
discloses that a topic of a meeting is compared to each attendee's personalized user
model to make recommendations.. Nowhere does Deng teach or suggest determining
whether a speaker is critical to the topic or removing content ... from summary in
response to determining that the speaker is not critical to the topic, as claimed. ”
Applicant’s arguments with respect to claim(s) 22 and 24 have been considered but are moot because the new ground of rejection does not rely on any reference applied in the prior rejection of record for any teaching or matter specifically challenged in the argument.
In response to the art rejection(s) of the remainder of dependent claims are rejected under 35 U.S.C 103, in case said claims are correspondingly discussed and/or argued for at least the same rationale presented in Remarks filed 10/22/2025, Examiner respectfully notes as follows. For completeness, should the mentioned claims be likewise traversed for similar reasons to independent claims 1, 8 and 15 correspondingly, Examiner respectfully directs Applicant to the same previous supra reasons provided in the response directed towards claims 1, 8 and 15 correspondingly discussed above. For at least the same supra provided reasons, Examiner likewise respectfully disagrees, and Applicant's arguments have been fully considered but they are not persuasive.
Claim Rejections - 35 USC § 103
The text of those sections of Title 35, U.S. Code not included in this action can be found in a prior Office action.
Claims 1, 4-8, 11-15 and 18-21, 25 are rejected under 35 U.S.C. 103 as being unpatentable over Nowak-Przygodzki et. al., US Patent 10,645,035(referred as Nowak) in view of Mody et al., US PgPub. 2020/0273453 further in view of Julia A Goldberg. Interrupting the discourse on interruptions: An analysis in terms of relationally neutral, power and rapport-oriented acts. Journal of Pragmatics, 14(6): 883–903, 1990 .
Regarding claim 1, Nowak teaches a computer-implemented machine learning method for generating real-time summaries, the method comprising: receiving a data during a conference session (see Nowak, col 16 lines 7-30 describes the conferencing setup and includes the data : the speech received and the agenda and calendar ); identifying a speech segment associated with the received data during [[a]] the conference session using a machine learning (ML) model (see Nowak, col 16 lines 31-52 discusses identifying the speech segment in conference mode of the automated assistant ( ML model) using the semantic processing and entity tagging ); generating a real-time transcript from the speech segment identified during the conference session using an automatic speech recognition (ASR) (see Nowak, col 19 lines 1-11 discusses speech to text processing on multiple distinct spoken utterances; Nowak, Fig. 1, 116 ( Text to speech module) in the Automated assistant 120); computing a lexical feature and a discourse feature from the real-time transcript(see Nowak, col 13 lines 49-61 discusses the semantic(lexical feature) processing of the utterances and checking with action items ( discourse features) of the agenda and updating accordingly); determining a drift in topic based on a pattern associated with the lexical feature or based on a distribution associated with the lexical feature and the discourse feature (see Nowak, col 13 lines 35-48 discusses topic classifier that identifies, from text generated from participant utterances, one or more topics that are raised and/or identifies when discussion has transitioned between different topics ; Nowak, col 15 lines 23-26, a new topic of discussion is detected based on the semantic processing to determine the context); determining a plurality of topics from the real-time transcript generated from the speech segment based on determining the drift in topic and further by using an ML classifier (see Nowak, col 13 lines 35-48 discusses the topic classifier based on the participant utterances using known techniques of topic classification, a topic classifier may employ a variety of known techniques of topic classification that are often used for document classification, such as expectation maximization, term-frequency-inverse document frequency (“TF-IDF”), na?ve Bayes classification, latent semantic indexing, support vector machines, artificial neural networks, decision trees, concept mining, etc.); generating a summary of the a topic of the plurality of topics (see Nowak, col 14 lines 9-20 generating a summary of the topics discussed ); and streaming the summary of the topic during the conference session (see Nowak, col 15 lines 38-44 discusses visual or audio output of the summary of the conference session. See Nowak, col 19 lines 48-50 discusses the meeting summary is generation).
Nowak teaches determining a plurality of topics from the real-time transcript generated from the speech segment based on determining the drift in topic and further by using an ML classifier; generating a summary of the a topic of the plurality of topics; and streaming the summary of the topic during the conference session, to further compact prosecution, Mody further teaches determining a plurality of topics from the real-time transcript generated from the speech segment based on determining the drift in topic (see Mody, [0040-0042] processes the identification of the topic and subtopics and using the HAC algorithm to cluster the vectors and determine the topics; Mody [0032] describes the processing of the similarity of the vectors and temporal distance ( determining the drift in topic) ); generating a summary of a topic of the plurality of topics (see Mody, [0034] the summary generation generates a meeting summary based on topics and subtopics, see Fig. 6B); and streaming the summary of the topic during the conference session (see Mody, [0045] output module provides the generated meeting summary; Mody Fig. 6B).
Nowak and Mody are considered to be analogous to the claimed invention because they relate to methods for automated meeting summarizations. Therefore, it would have been obvious to someone of ordinary skill in the art before the effective filing date of the claimed invention to have modified the teachings of Nowak on the method for identifying topics and generating meeting summary with automated topic identification and summarization and presentation teachings of Mody to improve providing insights into contents of the meeting( see Mody, [0002] ).
Nowak teaches computing a lexical feature and a discourse feature from the real-time transcript, however Nowak in view of Mody fails to teach wherein the discourse feature includes at least one or more interruptions in the speech segment.
However, Goldberg teaches computing a lexical feature and a discourse feature from the real-time transcript, wherein the discourse feature includes at least one or more types of interruptions in the speech segment, wherein the one or more types of interruptions include a neutral interruption, a power interruption associated with a decision being made, or a report interruption (see Goldberg, sect 3.1-3.2 The first step in the interpretive schema is to identify and separate relationally neutral interruptions from relationally loaded ones (i.e., those falling along the power-rapport continuum). Relationally neutral interruptions are those which address the immediate needs of the communicative situation. Neutral interruptions, in effect, initiate minor side, repair or misapprehension sequences; once completed the discourse is returned to its pre-interruption state thereby 'allowing' the interrupted speaker to continue where s/he left off. Power and rapport type interruptions are designed to satisfy li:;tener wants at the expense of his/her own obligations to support the rights (and wants) of the speaker to an unimpeded turn. Distinguishing between these two types of interruptions is the next step (step 2) of the heuristic. Power type interruptions are designed to wrest the discourse from the speaker by gaining control of the conversational process and/or content. Such power type interrupting the discourse type interruptions typically revolve topic change attempts accomplished by questions and requests (process control strategies) or by assertions or statements (content control strategies) whose proposition A content is unrelated to the specific topic at hand); determining a drift in topic based on a pattern associated with the lexical feature and the discourse feature or based on a distribution associated with the lexical feature and the discourse feature (see Goldberg, sect 3.2.1 Power type interruptions(discourse feature) are designed to wrest the discourse from the speaker by gaining control of the conversational process and/or content. Such power type interruptions typically revolve topic change attempts accomplished by questions and requests (process control strategies) or by assertions or statements (content control strategies) whose proposition A content is unrelated to the specific topic at hand(pattern associated with lexical features). When successful, content control interruptions constrain both the direction and substance of the discourse. Control is now in the hands of the interrupter (as in examples 9-12)). Examiner’s note : if the claims could further details on how the different interrupts are detected by the invention.
Nowak in view of Mody teach the method for identifying topics and generating meeting summary, however does not teach detection of different types of interruptions during a discussion. Goldberg teaches methods of identifying different types of interruptions during a conversation with examples. Using the known technique of interrupt detection as taught by Goldberg, to provide discourse feature from a conversation and change of topic in the references Nowak in view of Mody and to provide interruption type information, such as improved drift in topic would have been obvious to one of ordinary skill in the art.
Regarding claim 4, Nowak in view of Mody further in view of Goldberg teaches the computer-implemented machine learning method of claim 1. Mody further teaches generating the real-time transcript comprises tagging a speaker identity or a timestamp, and wherein generating the summary of the topic comprises generating the summary using the speaker identity or the timestamp (see Mody, [0040] discusses speaker diarization and time processing of the text. Mody, [0061] discusses the summary generation which includes additional metadata regarding the times at which each topic was discussed, the meeting participants that played major roles in discussing each topic, etc.).
Regarding claim 5, Nowak in view of Mody further in view of Goldberg teaches the computer-implemented machine learning method of claim 1. Nowak further teaches determining another topic from the real-time transcript generated from the speech segment(see Nowak, col 19 lines 21-27 discusses semantic processing the utterance(other topic) which is not relevant to the meeting discussion.. when the relevancy score fails a threshold ); determining an irrelevancy of the other topic (see Nowak, col 19 lines 21-27 if the semantic processing to produce the relevancy score which fails the threshold is interpreted as irrelevancy of the other topic ); and filtering out the other topic based on the irrelevancy (see Nowak, col 19 lines 21-34 discusses if the utterance fails the relevance threshold then no action is taken; filtering out the other topic).
Regarding claim 6, Nowak in view of Mody further in view of Goldberg teaches the computer-implemented machine learning method of claim 1. Mody further teaches generating the summary of the topic comprises generating an abstractive summary that includes independently chosen words from existing words in the real-time transcript (see Mody, [0070] discusses summary generated from the meeting; Mody, [0043] The resulting meeting summary may present a list of the topics discussed during the meeting along with statement determined to be relevant (e.g., have a high tf-idf value) to each topic.), and wherein streaming the summary of the topic comprises streaming the abstractive summary(see Modi, Fig 6A, [0070-0073] discusses the summary output and the topics and the playback of the portion of the recorded meeting via the user interface). Nowak further teaches generating the summary of the topic comprises generating an abstractive summary that includes independently chosen words from existing words in the real-time transcript, and wherein streaming the summary of the topic comprises streaming the abstractive summary (see Nowak, col 13 line 63- col 14 line 20 discusses generating a meeting summary based on the content of the discussion(real time transcript) learnt through semantic processing(independently chosen words) and processed to include topics discussed, action items created/addressed/modified, outcomes of the meeting and so forth and Nowak Fig. 2D, 2062, ).
Regarding claim 7, Nowak in view of Mody further in view of Goldberg teaches the computer-implemented machine learning method of claim 1. Mody further teaches processing the summary in response to generating the summary(see Mody, [0043] The summary generation module 210 generates a meeting summary based on the topics and subtopics identified by the topic identification module 208 for the videoconference meeting), wherein processing comprises adding a speaker identity or a timestamp(see Mody [0043] Further, the summary generation module 210 may include additional metadata regarding the times at which each topic was discussed, the meeting participants that played major roles in discussing each topic, etc. ); and wherein streaming the summary comprises streaming the summary in response to the processing(see Mody, [0043] The resulting meeting summary may present a list of the topics discussed during the meeting along with statement determined to be relevant to each topic. The topics may be presented according to the chronological order in which the topics were discussed during the meeting. Further, the topics may be presented along with data identifying times during the meeting at which each topic was presented. In some embodiments, the topics may be presented along with links that cause the recording of the meeting to be forwarded to the portion of the meeting during which the respective topic is discussed. A user may read the meeting summary to determine what was discussed as well as select to view any relevant portion ).
Regarding claim 8, is directed to a non-transitory computer-readable media claim corresponding to the method claim presented in claim 1 and is rejected under the same grounds stated above regarding claim 1.
Regarding claim 11, is directed to a non-transitory computer-readable media claim corresponding to the method claim presented in claim 4 and is rejected under the same grounds stated above regarding claim 4.
Regarding claim 12, is directed to a non-transitory computer-readable media claim corresponding to the method claim presented in claim 5 and is rejected under the same grounds stated above regarding claim 5.
Regarding claim 13, is directed to a non-transitory computer-readable media claim corresponding to the method claim presented in claim 6 and is rejected under the same grounds stated above regarding claim 6.
Regarding claim 14, is directed to a non-transitory computer-readable media claim corresponding to the method claim presented in claim 7 and is rejected under the same grounds stated above regarding claim 7.
Regarding claim 15, is directed to a machine learning system claim corresponding to the method claim presented in claim 1 and is rejected under the same grounds stated above regarding claim 1.
Regarding claim 18, is directed to a machine learning system claim corresponding to the method claim presented in claim 5 and is rejected under the same grounds stated above regarding claim 5.
Regarding claim 19, is directed to a machine learning system claim corresponding to the method claim presented in claim 6 and is rejected under the same grounds stated above regarding claim 6.
Regarding claim 20, is directed to a machine learning system claim corresponding to the method claim presented in claim 7 and is rejected under the same grounds stated above regarding claim 7.
Regarding claim 21, Nowak in view of Mody further in view of Goldberg teaches the computer-implemented machine learning method of claim 1. Mody further teaches further comprising generating a sentence vector for each sentence of the real-time transcript(see Mody, [0064] At 504, the vector clustering module 304 generates representative vectors for each statement), and wherein the determining the drift in topic is further based on the sentence vector(see Mody, [0066], [0066] At operation 506, the clustering module 306 clusters the representative vectors into vector clusters. Each resulting vector cluster includes a subset of the representative vectors that represent statements from the text that are likely part of the same topic ; Mody [0017, 0041, 0051] The cosine similarity value describes a measure of similarity between two representative vectors based on the angle between the two vectors. The temporal distance between two distances indicates an amount of time that elapsed between occurrence of the statements. The meeting summarization system uses the cosine similarity values and the temporal distances to determine vector distances between the representative vectors, which the HAC algorithm uses to cluster the representative vectors into vector clusters. Different topics to depict delineation one topic from another are interpreted as drift in topic).
Regarding claim 25, Nowak in view of Mody further in view of Goldberg teaches the computer-implemented machine learning method of claim 1. Mody further teaches wherein the summary is generated in response to a user selection of a participant of the conference session (see Mody, [0075] selection of the user interface element 624 corresponding to statement 1 causes presentation of the portion of the recorded meeting during which statement 1 was spoken by the meeting participants. Likewise, selection of the user interface element 626 corresponding to statement 2 causes presentation of the portion of the recorded meeting during which statement 2 was spoken by the meeting participants, Mody [0079] describes the updating of the summary section accordingly).
Claim 22 is rejected under 35 U.S.C. 103 as being unpatentable over Nowak-Przygodzki et. al., US Patent 10,645,035(referred as Nowak) in view of Mody et al., US PgPub. 2020/0273453 further in view of Julia A Goldberg. Interrupting the discourse on interruptions: An analysis in terms of relationally neutral, power and rapport-oriented acts. Journal of Pragmatics, 14(6): 883–903, 1990 further in view of Curtis et. al., US Patent 8,266,534.
Regarding claim 22, Nowak in view of Mody further in view of Goldberg teaches the computer-implemented machine learning method of claim 1. Nowak further teaches determining whether a speaker is critical to the topic (see Nowak, col 13 lines 1-10 Not every participant utterance is worthy of a response by automated assistant 120. For example, participants may engage in informal banter during the meeting to which they may not desire automated assistant 120 react( content not critical to the topic under discussion). Accordingly, in various implementations, automated assistant 120 may analyze various criteria to determine whether to inject into meetings content it retrieves based on semantic processing of the participants' discussions. In some implementations, automated assistant 120 may determine a relevancy score associated with information it obtains responsive to a participant's utterance); and removing content associated with the speaker from the summary in response to determining that the speaker is not critical to the topic (see Nowak, col 13 lines 15-20 On the other hand, if the retrieved information has a relevancy score that fails to satisfy such a threshold, automated assistant 120 may refrain from incorporating the information into the meeting discussion because that information may not likely be useful to, or well-received by, the participants). Nowak teaches removing content associated with the speaker from the summary in response to determining that the speaker is not critical to the topic based on the speaker content not relevant to the topic of discussion. To further teach determining whether a speaker is critical to the topic, Curtis teaches determining whether a speaker is critical to the topic (see Curtis, col 10 lines 8-17 discusses Filtering component 508 can identify a speaker, such as through voice-recognition or other manners of recognition, and filter out all the other speakers ( determining critical to the topic) ) and removing content associated with the speaker from the summary in response to determining that the speaker is not critical to the topic (see Curtis, 10 lines 17-22 Alternatively or additionally, filtering component 508 can remove information contributed by one or more identified people. As such, relevant information can be presented to a user while information not relevant or not desired can be disregarded and not output to the user. ). Examiner’s note : if the claims could further details on how the critical speaker is detected and how the non critical content is removed by the invention.
Nowak, Mody, Goldberg and Curtis are considered to be analogous to the claimed invention because they relate to methods for automating meetings. Therefore, it would have been obvious to someone of ordinary skill in the art before the effective filing date of the claimed invention to have modified the teachings of Nowak in view of Mody further in view of Goldberg on the method for identifying topics and generating meeting summary with the topic management teachings of Curtis to improve efficiency of meetings( see Curtis, col 1 lines 28-37).
Claims 23-24 are rejected under 35 U.S.C. 103 as being unpatentable over Nowak-Przygodzki et. al., US Patent 10,645,035(referred as Nowak) in view of Mody et al., US PgPub. 2020/0273453 further in view of Julia A Goldberg. Interrupting the discourse on interruptions: An analysis in terms of relationally neutral, power and rapport-oriented acts. Journal of Pragmatics, 14(6): 883–903, 1990 further in view of Wang et. al., US Patent 10,719,696 .
Regarding claim 23, Nowak in view of Mody further in view of Goldberg teaches the computer-implemented machine learning method of claim 1. However, Nowak in view of Mody further in view of Goldberg fails to teach determining sentiments and emotions associated with a speaker in the speech segment, wherein the summary of the topic is further based on the determined sentiments and emotions.
However, Wang teaches determining sentiments and emotions associated with a speaker in the speech segment, wherein the summary of the topic is further based on the determined sentiments and emotions(see Wang, col 3 lines 10-14 discusses determining emotions of participants to determine the interest of the topic; Wang, col 7 lines 8-10; (based on Specifications [0067]). Examiner’s note : if the claims could further details on how the sentiments and emotions speaker is processed with relevance to the topic by the invention.
Nowak, Mody, Goldberg and Wang are considered to be analogous to the claimed invention because they relate to methods for automating meetings. Therefore, it would have been obvious to someone of ordinary skill in the art before the effective filing date of the claimed invention to have modified the teachings of Nowak in view of Mody further in view of Goldberg on the method for identifying topics and generating meeting summary with the Interrelationships Among Participants And Topics In A Videoconferencing System teachings of Wang to improve flexibility in interaction among participants( see Wang, col 2 lines 55-59).
Regarding claim 24, Nowak in view of Mody further in view of Goldberg teaches the computer-implemented machine learning method of claim 1. However Nowak in view of Mody further in view of Goldberg fail to teach determining user engagements of the real-time transcript; and weighing topics of the plurality of topics based on the determined user engagements.
However, Wang teaches determining user engagements of the real-time transcript (see Wang, col 5 lines 24-26, col 5 lines 31-39 discusses collecting micro emotional feedback( user engagement)); and weighing topics of the plurality of topics based on the determined user engagements (see Wang, col 5 line 66- col 6 line 21, Fig. 6 discusses measures of interest in the topic of the participants and is relative scores are determined). The same motivation to combine as claim 23 applies here.
Claim 26 is rejected under 35 U.S.C. 103 as being unpatentable over Nowak-Przygodzki et. al., US Patent 10,645,035(referred as Nowak) in view of Mody et al., US PgPub. 2020/0273453 further in view of Julia A Goldberg. Interrupting the discourse on interruptions: An analysis in terms of relationally neutral, power and rapport-oriented acts. Journal of Pragmatics, 14(6): 883–903, 1990 further in view of Chen et. al. , US PgPub. 2016/0284354.
Regarding claim 26, Nowak in view of Mody further in view of Goldberg teaches the computer-implemented machine learning method of claim 1. Mody further teaches tagging a speaker in the conference session(see Mody, [0075] selection of the user interface element 624 corresponding to statement 1 causes presentation of the portion of the recorded meeting during which statement 1 was spoken by the meeting participants. Likewise, selection of the user interface element 626 corresponding to statement 2 causes presentation of the portion of the recorded meeting during which statement 2 was spoken by the meeting participants); and generating an abstractive summary associated with the speaker(see Mody, [0079] describes the updating of the summary section accordingly ). To further compact prosecution, Chen further teaches tagging a speaker in the conference session(see Chen, [0020] speech summarization program 118 may identify the voiceprint of the speaker from audio feed data received via network 108 (step 204)); and generating an abstractive summary associated with the speaker(see Chen, [0025] Speech summarization program 118 determines the key points made within the statements transcribed in step 210 (step 212). In the example embodiment, speech summarization program 118 determines key points by utilizing several methods, including monitoring for preselected keywords designated by participants or the host of the meeting [0026] Speech summarization program 118 generates and displays an overlay listing a speaker's statements that were determined key points in step 212 (step 214) ).
Nowak, Mody, Shires and Chen are considered to be analogous to the claimed invention because they relate to methods for automating conferences. Therefore, it would have been obvious to someone of ordinary skill in the art before the effective filing date of the claimed invention to have modified the teachings of Nowak in view of Mody further in view of Shires on the method for identifying topics and generating meeting summary with the speaker speech summarizations teachings of Gorman to improve remote conferencing ( see Gorman, [0002-0003]).
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
Krishnan et. al., US PgPub. 2022/0093093 teaches user interruption to determine change of topic (see Krishnan, [0350-0351]).
Bellamy et. al., US PgPub. 2017/0243171 teaches a method to list the critical speakers for the meeting and the difference determination circuit determines that when “Jon” and “Alan” speak that “who” is speaking is different than what is stated on the agenda. Thus, the difference determination circuit determines a difference from the agenda based on who is speaking (see Bellamy, Fig. 3, 4).
Banerjee, S., & Rudnicky, A. (2006), “A TextTiling based approach to topic boundary detection in meetings” discusses automatically detect boundaries between discussions of different topics in meetings using the TextTiling algorithm to the context of meetings.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to NANDINI SUBRAMANI whose telephone number is (571)272-3916. The examiner can normally be reached Monday - Friday 12:00pm - 5:00 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, Bhavesh M Mehta can be reached at (571)272-7453. 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.
/NANDINI SUBRAMANI/ Examiner, Art Unit 2656
/BHAVESH M MEHTA/ Supervisory Patent Examiner, Art Unit 2656