DETAILED ACTION
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Claim Interpretation
The following is a quotation of 35 U.S.C. 112(f):
(f) Element in Claim for a Combination. – An element in a claim for a combination may be expressed as a means or step for performing a specified function without the recital of structure, material, or acts in support thereof, and such claim shall be construed to cover the corresponding structure, material, or acts described in the specification and equivalents thereof.
The following is a quotation of pre-AIA 35 U.S.C. 112, sixth paragraph:
An element in a claim for a combination may be expressed as a means or step for performing a specified function without the recital of structure, material, or acts in support thereof, and such claim shall be construed to cover the corresponding structure, material, or acts described in the specification and equivalents thereof.
The claims in this application are given their broadest reasonable interpretation using the plain meaning of the claim language in light of the specification as it would be understood by one of ordinary skill in the art. The broadest reasonable interpretation of a claim element (also commonly referred to as a claim limitation) is limited by the description in the specification when 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, is invoked.
As explained in MPEP § 2181, subsection I, claim limitations that meet the following three-prong test will be interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph:
(A) the claim limitation uses the term “means” or “step” or a term used as a substitute for “means” that is a generic placeholder (also called a nonce term or a non-structural term having no specific structural meaning) for performing the claimed function;
(B) the term “means” or “step” or the generic placeholder is modified by functional language, typically, but not always linked by the transition word “for” (e.g., “means for”) or another linking word or phrase, such as “configured to” or “so that”; and
(C) the term “means” or “step” or the generic placeholder is not modified by sufficient structure, material, or acts for performing the claimed function.
Use of the word “means” (or “step”) in a claim with functional language creates a rebuttable presumption that the claim limitation is to be treated in accordance with 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph. The presumption that the claim limitation is interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, is rebutted when the claim limitation recites sufficient structure, material, or acts to entirely perform the recited function.
Absence of the word “means” (or “step”) in a claim creates a rebuttable presumption that the claim limitation is not to be treated in accordance with 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph. The presumption that the claim limitation is not interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, is rebutted when the claim limitation recites function without reciting sufficient structure, material or acts to entirely perform the recited function.
Claim limitations in this application that use the word “means” (or “step”) are being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, except as otherwise indicated in an Office action. Conversely, claim limitations in this application that do not use the word “means” (or “step”) are not being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, except as otherwise indicated in an Office action.
This application includes one or more claim limitations that do not use the word “means,” but are nonetheless being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, because the claim limitation(s) uses a generic placeholder that is coupled with functional language without reciting sufficient structure to perform the recited function and the generic placeholder is not preceded by a structural modifier. Such claim limitations are: “a text preprocessing unit”, “a text markup unit”, “a contextual window control unit”, “a sentiment classification unit”, “an emotional text markup unit” and “an avatar generation unit” in claims 11-20.
Because these claim limitations are being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, they are being interpreted to cover the corresponding structure described in the specification as performing the claimed function, and equivalents thereof.
If applicant does not intend to have these limitations interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, applicant may: (1) amend the claim limitation(s) to avoid it/them being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph (e.g., by reciting sufficient structure to perform the claimed function); or (2) present a sufficient showing that the claim limitation(s) recite(s) sufficient structure to perform the claimed function so as to avoid it/them being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph.
Claim Rejections - 35 USC § 112
The following is a quotation of 35 U.S.C. 112(b):
(b) CONCLUSION.—The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the inventor or a joint inventor regards as the invention.
The following is a quotation of 35 U.S.C. 112 (pre-AIA ), second paragraph:
The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the applicant regards as his invention.
Claims 9, 14 and 15 rejected under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), second paragraph, as being indefinite for failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor (or for applications subject to pre-AIA 35 U.S.C. 112, the applicant), regards as the invention.
Claim 9 recites the limitation "the sentiment analysis model" in line 1. There is insufficient antecedent basis for this limitation in the claim.
Claim 14 recites the limitation "the emotional analysis machine learning model" in line 2. There is insufficient antecedent basis for this limitation in the claim.
Claim 15 recites the limitation "the emotional analysis model" in line 1. There is insufficient antecedent basis for this limitation in the claim.
Claim Rejections - 35 USC § 103
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
Claims 1, 4, 5, 8-12, 14-16, 19 and 20 are rejected under 35 U.S.C. 103 as being unpatentable over Jawal et al (US Patent 11,580,961; hereinafter “Jawal”) in view of Asi (US PG Pub 20190362713).
As per claim 1, Jawal discloses: A method for automated emotional text analysis and markup, comprising: receiving input text data (Jawal; Fig. 11, item 1102; Col. 30, lines 29-38 - In this example, for some embodiments, the conversation and the trackers may be considered text-based. After a start block, at block 1102, in one or more of the various embodiments, a conversation stream may be provided to a tracker engine. As described above, a conversation stream may be provided to a tracker engine. As described above, in some embodiments, audio (e.g., speech) conversation streams may be pre-processed into text comprising individual words that may be provided as a stream of text words); preprocessing said input text data to identify and extract text segments (Jawale; Col. 30, lines 35-50 - audio (e.g., speech) conversation streams may be pre-processed into text comprising individual words that may be provided as a stream of text words. In one or more of the various embodiments, tracker engines may be arranged to buffer one or more individual words (identify and extract text segments) before evaluating them with one or more tracker models. In one or more of the various embodiments, the buffers may be FIFO (first in first out) queues, or the like, that may collect a portion of the conversation stream in the same order the words are spoken or written (in the case of chats or emails)); applying a sliding window mechanism to each text segment to create a context window (Jawale; Col. 30, lines 39-60 - One of ordinary skill in the art will appreciate that these buffers may be referred to as sliding windows because they hold a range of words from the conversation. At decision block 1104, in one or more of the various embodiments, if the sliding window of one or more tracker models may be fit or filled, control may flow to block 1106; otherwise, control may loop back to decision block 1104) for sentiment analysis (Jawale; Col. 4, lines 45-54 - tracker models may be employed to determine in real-time if the speech in a conversation that may match the meaning, usage, or sentiment of a corresponding tracker vocabulary); classifying a sentiment of the text within the context window using an emotional analysis model (Jawale; Col. 4, lines 45-54 - tracker models may be employed to determine in real-time if the speech in a conversation that may match the meaning, usage, or sentiment of a corresponding tracker vocabulary; see also Fig. 10, Col. 27, lines 55-67 & Col. 28, lines 1-4 – generating tracker metrics using tracker vocabularies that represent sentiment), wherein an input of the emotional analysis model is a context window and a result of the classification is a verdict containing at least one sentiment class with an accuracy value, wherein the accuracy value characterizes the confidence level of the classification by indicating the probabilistic likelihood of the text segment and the sentiment class (Jawale; Col. 24. Lines 17-27 - In some embodiments, tracker engines may be arranged to generate match score (accuracy score), such as, match score 710 that indicates if the words/phrases in the sliding window matched the tracker model. In some embodiments, the match score may be real number value representing the closeness or confidence the match. Also, in some embodiments, the match score may be discrete values representing match or non-match. Further, in some embodiments, tracker engines may be arranged to provide a match score that includes more than one value or component, including match/non-match, confidence score, or the like); extending the window size for the text segment to create an extended window if the classification accuracy is below a predefined threshold (Jawale; Col. 31, lines 11-14 - Also, in some embodiments, tracker engines may be configured initialize sliding window sizes to a lower value (e.g., one word) and increment the size until match rates are above a defined threshold value; see also Col. 23, lines 56-66 - …For example, in some embodiments, tracker engines may be configured to progressively adjust sliding window sizes based on match results (create an extended window if the classification accuracy is below a predefined threshold). For example, a tracker engine may be configured automatically increase or decrease the sliding window size to identify sizes that may be more effective or efficient for a given tracker model. Also, for example, a sliding window may initially be set to one word, then the tracker engine may increment the size of the sliding window, monitoring the effect on match results) and classifying the sentiment of the text segment within the extended window (Jawale; Col. 4, lines 45-54 - tracker models may be employed to determine in real-time if the speech in a conversation that may match the meaning, usage, or sentiment of a corresponding tracker vocabulary; see also Fig. 10, Col. 27, lines 55-67 & Col. 28, lines 1-4 – generating tracker metrics using tracker vocabularies that represent sentiment); associating the classification verdict with the respective text segments to generate sentiment-classified text segments (Jawale; Col. 24. Lines 17-27 - In some embodiments, tracker engines may be arranged to generate match score, such as, match score 710 that indicates if the words/phrases in the sliding window matched the tracker model. In some embodiments, the match score may be real number value representing the closeness or confidence the match. Also, in some embodiments, the match score may be discrete values representing match or non-match (associating). Further, in some embodiments, tracker engines may be arranged to provide a match score that includes more than one value or component, including match/non-match, confidence score, or the like); and generating media content based on the sentiment-classified text segments (Jawale; Col. 19, lines 13-21 - a speech analysis engine may be configured to automatically generate notifications or events if portions of a conversation in a conversation stream match one or more tracker models. Likewise, for example, tracker engines may be arranged to generate reports in the form of log files, user interfaces, dashboards, or the like (generating media content), that may provide real-time feedback regarding the occurrence of speech that matches one or more tracker models).
Although Jawale disclose the use of tracker models to determine sentiment, Jawale fails to explicitly disclose that the emotional analysis model is a machine-learning model, wherein an input of the emotional analysis machine-learning model is a context window and a result of the classification is a verdict containing at least one sentiment class with an accuracy value, wherein the accuracy value characterizes the confidence level of the classification by indicating the probabilistic likelihood of the text segment and the sentiment class. Asi does teach that the emotional analysis model is a machine-learning model, wherein an input of the emotional analysis machine-learning model is a context window and a result of the classification is a verdict containing at least one sentiment class with an accuracy value, wherein the accuracy value characterizes the confidence level of the classification by indicating the probabilistic likelihood of the text segment and the sentiment class (Asi; p. 0037 - it is contemplated that the token group analyzing component 240 has been trained with training data that enables the neural network (emotional analysis machine-learning model) component 150 to determine, with a calculated level of confidence determined by the neural network component 150, an emotional score (e.g., a confidence or likelihood) that input data being analyzed corresponds to one of a plurality of determinable human emotions. In various embodiments, an emotional score can be calculated by emotion scoring component 242, which calculates a probability value for each emotion in a set of determinable human emotions based on the input data provided thereto; see also p. 0021 & 0025). Therefore, it would have been obvious to one of ordinary skill in the art, to modify the method of Jawal to include that the emotional analysis model is a machine-learning model, wherein an input of the emotional analysis machine-learning model is a context window and a result of the classification is a verdict containing at least one sentiment class with an accuracy value, wherein the accuracy value characterizes the confidence level of the classification by indicating the probabilistic likelihood of the text segment and the sentiment class, as taught by Asi, so that an ideal and improved technique would analyze more than sentences, but variable blocks of sentences, to determine whether the author's emotion is consistent as sentences flow from one to the next… The described embodiments provide improved results in the field of textual analytics by recognizing a deeper level of sentiment in text-based content, that being emotion, and further recognizing shifts in emotions so that portions of an electronic document can be easily identified based on a determined related emotion (Asi; p. 0015-0016).
As per claim 4, Jawal in view of Asi disclose:
The method of claim 1, further comprising: determining a textual domain of the input text data (Jawal; Col. 17, lines 22-48 - one or more tracker vocabularies, such as, tracker vocabulary 406 may be provided to tracker engine 402… tracker vocabularies may correspond set of concepts or topics (textual domain) that a tracker engine may track within a conversation stream. Accordingly, in some embodiments, tracker vocabularies may be comprised of words or phrases that may be related to particular concepts, semantics, or topics of interest); and applying the sliding window mechanism for each text segment based on the textual domain (Jawale; Col. 30, lines 39-60 - One of ordinary skill in the art will appreciate that these buffers may be referred to as sliding windows because they hold a range of words from the conversation. At decision block 1104, in one or more of the various embodiments, if the sliding window of one or more tracker models may be fit or filled, control may flow to block 1106; otherwise, control may loop back to decision block 1104).
As per claim 5, Jawal in view of Asi disclose:
The method of claim 4, wherein the emotional analysis model is selected based on the textual domain of the input text data (Jawal; Col. 17, lines 55-67 & Col. 18, lines 1-11 - tracker engines may be arranged to provide tracker vocabularies to a universal generalization model, such as, universal generalization model 410 to generate one or more tracker models… universal generalization models may be machine learning models that may be trained to predict a generalized vocabulary from the specific samples included in tracker vocabularies… Thus, the tracker vocabularies, such as used to determine the textual domain, are used in the selection and/or generation of the analysis model such as the tracker model used to determine the sentiment).
As per claim 8, Jawal in view of Asi discloses:
The method of claim 1, wherein the text segment is a word, a phrase, or a sentence (Jawal; Col. 25, lines 25-29 - analysis engines or tracker engines may be arranged to process text-based conversations word-by-word, sentence-by-sentence, paragraph-by-paragraph (snippets)).
As per claim 9, Jawal in view of Asi disclose:
The method of claim 1, wherein the sentiment analysis model is a support vector machine, a deep neural network, or a recurrent neural network (Jawal; Col. 22, lines 28-31 - universal generalization model 606 may be a deep learning artificial neural network trained to predict a generalization vocabulary for a given tracker vocabulary).
As per claim 10, Jawal in view of Asi disclose:
The method of claim 1, upon which claim 10 depends. Asi teaches compiling all sentiment-classified text segments and associated emotional tags to form a marked-up text for generation of media content (Asi; p. 0042 - The neural network component 150 can also include a token group tagging component 260 that generates a tag, index entry, or other metadata for association with the content included in the selected token group (e.g., the representative token group). In various embodiments, the tag can be a comment, an annotation, a highlight with a color corresponding to an emotion, or any other visual indicator that can be embedded into the electronic document such that when the electronic document is provided for display via a compatible application, the visual indicator corresponding to the content included in the selected token group is presented with the content). Therefore, it would have been obvious to one of ordinary skill in the art, to modify the method of Jawal to include compiling all sentiment-classified text segments and associated emotional tags to form a marked-up text for generation of media content, as taught by Asi, so that an ideal and improved technique would analyze more than sentences, but variable blocks of sentences, to determine whether the author's emotion is consistent as sentences flow from one to the next… The described embodiments provide improved results in the field of textual analytics by recognizing a deeper level of sentiment in text-based content, that being emotion, and further recognizing shifts in emotions so that portions of an electronic document can be easily identified based on a determined related emotion (Asi; p. 0015-0016).
As per claim 11, Jawal discloses:
A system for automated emotional text analysis and markup, comprising: at least one processor (Jawal; Fig. 3, item 302) and memory (Jawal; Fig. 3, item 304) operably coupled to the at least one processor (Jawal; Col. 13, lines 30-32 - network computer 300 may include a processor 302 that may be in communication with a memory 304 via a bus 328); instructions that, when executed (Col. 15, lines 8-11 - Memory 304 illustrates an example of computer-readable storage media (devices) for storage of information such as computer-readable instructions), cause the at least one processor to implement: a text preprocessing unit configured to receive input text data and parse said data into text segments (Jawale; Col. 30, lines 35-50 - audio (e.g., speech) conversation streams may be pre-processed into text comprising individual words that may be provided as a stream of text words. In one or more of the various embodiments, tracker engines may be arranged to buffer one or more individual words before evaluating them with one or more tracker models. In one or more of the various embodiments, the buffers may be FIFO (first in first out) queues, or the like, that may collect a portion of the conversation stream in the same order the words are spoken or written (in the case of chats or emails) (parse said data into text segments)); a text markup unit configured to apply contextual analysis to the text segments (Jawale; Col. 30, lines 39-60; see also Col. 4, lines 45-54; see also Fig. 10, Col. 27, lines 55-67 & Col. 28, lines 1-4), the text markup unit comprising: a contextual window control unit configured to define context windows (Jawale; Col. 30, lines 39-60 - One of ordinary skill in the art will appreciate that these buffers may be referred to as sliding windows because they hold a range of words from the conversation. At decision block 1104, in one or more of the various embodiments, if the sliding window of one or more tracker models may be fit or filled, control may flow to block 1106; otherwise, control may loop back to decision block 1104) for sentiment analysis on the text segments (Jawale; Col. 4, lines 45-54 - tracker models may be employed to determine in real-time if the speech in a conversation that may match the meaning, usage, or sentiment of a corresponding tracker vocabulary), a sentiment classification unit configured to classify a sentiment of the text within the context windows (Jawale; Col. 4, lines 45-54 - tracker models may be employed to determine in real-time if the speech in a conversation that may match the meaning, usage, or sentiment of a corresponding tracker vocabulary; see also Fig. 10, Col. 27, lines 55-67 & Col. 28, lines 1-4 – generating tracker metrics using tracker vocabularies that represent sentiment); and an avatar generation unit configured to generate media content based on the sentiment-classified and emotionally annotated text segments (Jawale; Col. 19, lines 13-21 - a speech analysis engine may be configured to automatically generate notifications or events if portions of a conversation in a conversation stream match one or more tracker models. Likewise, for example, tracker engines may be arranged to generate reports in the form of log files, user interfaces, dashboards, or the like (generating media content), that may provide real-time feedback regarding the occurrence of speech that matches one or more tracker models). Jawale, however, fails to explicitly disclose an emotional text markup unit configured to annotate the text segments with emotional tags based on the sentiment classification. Asi does teach an emotional text markup unit configured to annotate the text segments with emotional tags based on the sentiment classification (Asi; p. 0030 - …the neural network component 115 can perform operations disclosed herein to tag consecutive portions of content within the electronic document 115 to indicate a corresponding determined emotion…). Therefore, it would have been obvious to one of ordinary skill in the art to modify the system of Jawale to include an emotional text markup unit configured to annotate the text segments with emotional tags based on the sentiment classification, as taught by Asi, so that an ideal and improved technique would analyze more than sentences, but variable blocks of sentences, to determine whether the author's emotion is consistent as sentences flow from one to the next… The described embodiments provide improved results in the field of textual analytics by recognizing a deeper level of sentiment in text-based content, that being emotion, and further recognizing shifts in emotions so that portions of an electronic document can be easily identified based on a determined related emotion (Asi; p. 0015-0016).
As per claim 12, Jawal in view of Asi disclose: The system of claim 11, upon which claim 12 depends. And further, Asi teaches wherein the sentiment classification unit further comprises an emotional analysis machine learning model for classifying text segments (Asi; p. 0037 - it is contemplated that the token group analyzing component 240 has been trained with training data that enables the neural network (emotional analysis machine-learning model) component 150 to determine, with a calculated level of confidence determined by the neural network component 150, an emotional score (e.g., a confidence or likelihood) that input data being analyzed corresponds to one of a plurality of determinable human emotions. In various embodiments, an emotional score can be calculated by emotion scoring component 242, which calculates a probability value for each emotion in a set of determinable human emotions based on the input data provided thereto; see also p. 0021 & 0025). Therefore, it would have been obvious to one of ordinary skill in the art, to modify the method of Jawal to include wherein the sentiment classification unit further comprises an emotional analysis machine learning model for classifying text segments, as taught by Asi, so that an ideal and improved technique would analyze more than sentences, but variable blocks of sentences, to determine whether the author's emotion is consistent as sentences flow from one to the next… The described embodiments provide improved results in the field of textual analytics by recognizing a deeper level of sentiment in text-based content, that being emotion, and further recognizing shifts in emotions so that portions of an electronic document can be easily identified based on a determined related emotion (Asi; p. 0015-0016).
As per claim 14, Jawal in view of Asi disclose:
The system of claim 11, wherein the text preprocessing unit is further configured to identify a textual domain of the input text data (Jawal; Col. 17, lines 22-48 - one or more tracker vocabularies, such as, tracker vocabulary 406 may be provided to tracker engine 402… tracker vocabularies may correspond set of concepts or topics (textual domain) that a tracker engine may track within a conversation stream. Accordingly, in some embodiments, tracker vocabularies may be comprised of words or phrases that may be related to particular concepts, semantics, or topics of interest), wherein the emotional analysis machine learning model is particularly trained for the textual domain (Jawal; Col. 17, lines 64-67 - universal generalization models may be machine learning models that may be trained to predict a generalized vocabulary from the specific samples included in tracker vocabularies).
As per claim 15, Jawal in view of Asi disclose the system of claim 11. Claim 15 particularly contains language that is similar to the language of claim 9 and, therefore, is rejected similarly.
As per claim 16, Jawal in view of Asi disclose the system of claim 11. Claim 16 particularly contains language that is similar to the language of claim 10 and, therefore, is rejected similarly.
As per claim 19, Jawal in view of Asi disclose: The system of claim 11, wherein the text markup unit is further configured to reprocess the text segments with extended context window, when initial sentiment classification of the context window does not meet a predefined accuracy threshold (Jawale; Col. 31, lines 11-14 - Also, in some embodiments, tracker engines may be configured initialize sliding window sizes to a lower value (e.g., one word) and increment the size until match rates are above a defined threshold value; see also Col. 23, lines 56-66 - …For example, in some embodiments, tracker engines may be configured to progressively adjust sliding window sizes based on match results (create an extended window if the classification accuracy is below a predefined threshold). For example, a tracker engine may be configured automatically increase or decrease the sliding window size to identify sizes that may be more effective or efficient for a given tracker model. Also, for example, a sliding window may initially be set to one word, then the tracker engine may increment the size of the sliding window, monitoring the effect on match results). As per claim 20, Jawal in view of Asi disclose the system of claim 11. Claim 20 particularly contains language that is similar to the language of claim 8 and, therefore, is rejected similarly.
Claims 2, 3 and 13 are rejected under 35 U.S.C. 103 as being unpatentable over Jawal in view of Asi and in view of Mishra (US PG Pub 20190172458).
As per claim 2, Jawal in view of Asi disclose:
The method of claim 1, upon which claim 2 depends. Jawal in view of Asi, however, fail to teach determining a language of the input text data; and applying the sliding window mechanism for each text segment based on the language. Mishra does teach determining a language of the input text data (Mishra; Fig. 2, item 240; p. 0041 - extracting contextual information 240 from neighboring segments… The contextual information can include data and estimations about the speaker such as gender, age, native language, etc., and fusion rules, window size, and so on); and applying the sliding window mechanism for each text segment based on the language (Mishra; p. 0037 - The contextual information can include data and estimations about the speaker such as gender, age, native language, etc., and fusion rules, window size, and so on. In embodiments, the successive, overlapped speech segments are windowed around 1200 ms. The window sizes can be varied to improve accuracy, to adjust computational complexity, and so on). Therefore, it would have been obvious to one of ordinary skill in the art to modify the method of Jawal in view of Asi to include determining a language of the input text data; and applying the sliding window mechanism for each text segment based on the language, as taught by Mishra, so that differences in language formality and idiomatic expressions across such groups and cultures can be considered for the emotional state analysis of a user (Mishra; p. 0031).
As per claim 3, Jawal in view of Asi and Mishra disclose:
The method of claim 2, upon which claim 3 depends. And further, Mishra teaches wherein the emotional analysis model is selected based on the language of the input text data (Mishra; p. 0031 - Training for cross-language speech analysis can include training data across language groups and across different cultures that use those language groups… The “emotional analysis model” selection is provided by the training of speech analysis to detect emotional states across multiple languages).
Therefore, it would have been obvious to one of ordinary skill in the art to modify the method of Jawal in view of Asi to include wherein the emotional analysis model is selected based on the language of the input text data, as taught by Mishra, so that differences in language formality and idiomatic expressions across such groups and cultures can be considered for the emotional state analysis of a user (Mishra; p. 0031).
As per claim 13, Jawal in view of Asi disclose:
The system of claim 12, upon which claim 13 depends. Jawal in view of Asi, however, fail to teach wherein the text preprocessing unit is further configured to determine a language of the input text data, wherein the emotional analysis machine learning model is particularly trained for the language. Mishra does teach wherein the text preprocessing unit is further configured to determine a language of the input text data (Mishra; Fig. 2, item 240; p. 0041 - extracting contextual information 240 from neighboring segments… The contextual information can include data and estimations about the speaker such as gender, age, native language, etc., and fusion rules, window size, and so on), wherein the emotional analysis machine learning model is particularly trained for the language (Mishra; p. 0031 - Training for cross-language speech analysis can include training data across language groups and across different cultures that use those language groups). Therefore, it would have been obvious to one of ordinary skill in the art to modify the method of Jawal in view of Asi to include wherein the text preprocessing unit is further configured to determine a language of the input text data, wherein the emotional analysis machine learning model is particularly trained for the language, as taught by Mishra, so that differences in language formality and idiomatic expressions across such groups and cultures can be considered for the emotional state analysis of a user (Mishra; p. 0031).
Claims 6, 7, 17 and 18 are rejected under 35 U.S.C. 103 as being unpatentable over Jawal in view of Asi and in view of Coursey et al. (US Patent 11,645,479; hereinafter “Coursey”).
As per claim 6, Jawal in view of Asi disclose:
The method of claim 1, upon which claim 6 depends. Jawal in view of Asi, however, fail to teach wherein the output of the sentiment classification is used as an input for a speech generator to create audio content that reflects the tagged emotional states. Coursey does teach wherein the output of the sentiment classification is used as an input for a speech generator to create audio content that reflects the tagged emotional states (Coursey; Col. 7, lines 55-67 & Col. 8, lines 1-37 - …abstracted tree search process (of which MCTS is a variant) which utilizes the contextual descriptions in a deliberative search process to render and select a next action. or utterance for speech, based on the projected expected outcomes as determined by the search process… The contextual description may include information regarding the selected personality and motivation of the virtual agent being simulated, as well as selected parameters and perceived parameters regarding the personality and motivation of the other various participants to the conversation exchange. For example, the virtual agent may be defined as being helpful, cheery and talkative. Such adjectives will influence the language model generation profile, which in turn influences the sentiment of text associated with the given character…). Therefore, it would have been obvious to one of ordinary skill in the art to modify the method of Jawal in view of Asi to include wherein the output of the sentiment classification is used as an input for a speech generator to create audio content that reflects the tagged emotional states, as taught by Coursey, because language models can generate many plausible responses, but can lack a way of selecting the “best” response versus the “most probable” response. The tree search process selects the “best” response leading to the highest average expected outcome based on sample-based projections of the future (Coursey; Col. 2, lines 31-37).
As per claim 7, Jawal in view of Asi discloses:
The method of claim 1, upon which claim 7 depends. Jawal in view of Asi, however, fail to teach wherein the output of the sentiment classification is used as an input for generating digital avatar movement. Coursey does teach wherein the output of the sentiment classification is used as an input for generating digital avatar movement (Coursey; Col. 3, lines 62-65 - In Step 007, the action selected in Step 005 is implemented through interpretation as either a user interaction, animation action (digital avatar movement) or as a system internal action; see also claim 8 - wherein in addition to textual expression the virtual language agent responds with animation and virtual effectors). Therefore, it would have been obvious to one of ordinary skill in the art to modify the method of Jawal in view of Asi to include wherein the output of the sentiment classification is used as an input for generating digital avatar movement, as taught by Coursey, because language models can generate many plausible responses, but can lack a way of selecting the “best” response versus the “most probable” response. The tree search process selects the “best” response leading to the highest average expected outcome based on sample-based projections of the future (Coursey; Col. 2, lines 31-37).
As per claim 17, Jawal in view of Asi disclose the system of claim 11. Claim 17 particularly contains language that is similar to the language of claim 6 and, therefore, is rejected similarly.
As per claim 18, Jawal in view of Asi disclose the system of claim 16. Claim 18 particularly contains language that is similar to the language of claim 7 and, therefore, is rejected similarly.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure. The prior art made of record and not relied upon includes:
M. A. Masood et al., "Context-Aware Sliding Window for Sentiment Classification," in IEEE Access, vol. 8, pp. 4870-4884, 2020. This non-patent literature discloses: In this research, we hypothesize that the past sentiments help the classifier to effectively link the user’s history along with the contents of the current tweet. Thus, allowing learning algorithms to correlate past activities in determining the current sentiments. For this sake, we propose three sliding window features to accumulate past sentiments from the time series data. In this paper, we propose seven variations of Context-aware Sliding Window (CSW) features on different machine learning and deep learning algorithms. Furthermore, we propose a temporal dataset of user tweets, which is manually labeled by nine human annotators. The proposed dataset consists of 36 users having 4,557 tweets. Results indicate significant improvements over six state-of-the-art baseline methods (Abstract).
Any inquiry concerning this communication or earlier communications from the examiner should be directed to Rodrigo A Chavez whose telephone number is (571)270-0139. The examiner can normally be reached Monday - Friday 9-6 ET.
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, Richemond Dorvil can be reached at 5712727602. 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.
/RODRIGO A CHAVEZ/Examiner, Art Unit 2658
/RICHEMOND DORVIL/Supervisory Patent Examiner, Art Unit 2658
Prosecution Insights