System and method for temporal interpretation of multimodal communication patterns

Abstract

A system and method for interpreting communication patterns from a user is disclosed. The system comprises a processor configured to receive at least one type of communication and determine at least one communication pattern. The communication pattern includes a tone, rhythm, speed, discontinuity, physiological pattern, or any combination thereof. The processor determines at least one metaphorical representation based on the determined communication patterns and determines at least one interpretation of the received communication based on the determined metaphorical representations and the determined communication patterns. The processor is further configured to perform at least one action based on the at least one determined interpretation.

Description

SYSTEM AND METHOD FOR TEMPORAL INTERPRETATION OF MULTIMODAL COMMUNICATION PATTERNSFIELD OF INVENTION

[0001] The present disclosure relates in general to interpreting communication patterns. More particularly, the present disclosure relates to a system and method for interpreting multimodal communication patterns based on verbal or non-verbal cues.BACKGROUND OF THE INVENTION

[0002] Conventional communication-analysis systems rely on semantic interpretation, facial cues, acoustic sentiment, or behavioral prediction to infer a user’s state. These approaches operate primarily on consciously modifiable signals and therefore provide limited insight into the underlying timing dynamics of multimodal communication. Existing systems are further constrained by their reliance on surface-level patterns, which may not accurately reflect the early autonomic changes that precede conscious interpretation. As a result, responses generated by such systems may not correspond to the user’s pre-cognitive physiological state. There is therefore a need for a system that interprets multimodal communication patterns using timing-based alignment across verbal and non-verbal channels, enabling detection of pre-cognitive rhythmic changes that occur prior to semantic processing.SUMMARY OF THE INVENTION

[0003] In an aspect of the present disclosure, a system and method for interpreting at least one type of multimodal communication using an artificial intelligence (Al) based architecture is disclosed. The system includes at least one processor and a memory storing instructions executable by the processor. The processor is configured to receive the at least one type of communication and determine at least one communication timing pattern associated with the at least one received type ofcommunication without semantic or contextual analysis of the at least one received type of communication.

[0004] The at least one communication timing pattern may include a communication tone, a communication rhythm, a communication pacing, at least one communication pause, at least one communication discontinuity, at least one micro-timing variation, at least one physiological timing cue, or any combination thereof associated with the at least one received type of communication.

[0005] Thereafter, the processor is configured to determine at least one representational mapping (e.g., at least one rhythmic category, symbolic template, timing-based pattern structure, or metaphorical representation) corresponding to the at least one determined communication timing pattern. Based on the at least one determined communication timing pattern and the at least one representational mapping, the processor determines at least one interpretation of the at least one received type of communication and performs at least one action based on the at least one determined interpretation.

[0006] In another aspect, the processor may include an Al module trained to analyze at least one multimodal timing signal associated with the at least one received type of communication.

[0007] In an aspect, the at least one communication timing pattern may further include a communication style, a communication method, a communication micropause, a communication inconsistency, a physical manner of communication, at least one timing-based content characteristic, or any combination thereof. The at least one timing-based content characteristic may comprise a frequency, variation, timing, interval, or timespan associated with the usage of at least one character, word, gesture, action, or any combination thereof in the at least one received type of communication.

[0008] In another aspect, when the at least one received type of communication includes at least one content-bearing signal, the processor is configured to identify at least one time-delay marker associated with the communication tone, the communication rhythm, the communication discontinuity, the communication pause, the communication inconsistency, or any combination thereof. Theprocessor may assign at least one weighted priority to the at least one identified time-delay marker. The processor determines the at least one interpretation of the at least one received type of communication based on the at least one identified time-delay marker and the at least one assigned weighted priority.

[0009] Based on the at least one assigned weighted priority, the processor may generate at least one output indicative of or associated with the at least one determined interpretation.

[0010] In another aspect, the at least one action may correspond to generating at least one output indicative of or based on the at least one determined interpretation. The at least one output may include at least one text response, at least one audio response, at least one visual indication, at least one haptic feedback, or any combination thereof. The at least one action may further correspond to modulating the at least one generated output based on the at least one determined interpretation. The modulation includes introducing at least one time delay at a beginning, during, or after providing the at least one generated output.[OH] In another aspect, the at least one received type of communication may include at least one of a text communication, a video communication, an audio communication, a haptic communication, a virtual -reality communication, an augmented-reality communication, or any combination thereof.

[0012] In yet another aspect, the processor may map the at least one determined communication timing pattern to a plurality of predefined timing-based categories. The plurality of predefined timing-based categories may be associated with a plurality of predefined representational templates. The processor may determine the at least one representational mapping of the at least one determined communication timing pattern based on the mapping.

[0013] In another aspect, the processor may determine at least one somatic marker associated with the at least one received type of communication based on the at least one determined communication timing pattern, the at least one determined representational mapping, a content included in the at least one received communication, or any combination thereof. Based on the at least one somatic marker, the processor may recommend at least one physiological actioncorresponding to the at least one determined representational mapping, the at least one determined interpretation, or any combination thereof.

[0014] In another aspect, the at least one representational mapping may include at least one metaphorical symbol, abstract template, timing-based structural pattern, sound, image, or any combination thereof.

[0015] In another aspect, the at least one generated output may include a plurality of output components, and the processor may provide the plurality of output components at one or more predefined intervals.

[0016] In another aspect, the processor may assign at least one communication tone, communication rhythm, communication pacing, communication pause, communication discontinuity, communication inconsistency, or physical manner of communication to at least one generated output based on the at least one determined representational mapping, the at least one determined interpretation, or any combination thereof.

[0017] In another aspect, the processor may determine a communication language, a communication environment, a communication location, or any combination thereof based on the at least one received type of communication. The at least one generated interpretation may be adapted to the at least one determined language, environment, or location.

[0018] In another aspect, the system may optionally receive physiological data from at least one sensor associated with at least one user and may generate the at least one output based on a combination of the multimodal timing signals and the at least one determined interpretation.

[0019] In another aspect, the processor may predict at least one user attribute, including an estimated age of the user, based on the at least one received type of communication, the at least one determined communication timing pattern, the at least one determined interpretation, or any combination thereof.

[0020] In another aspect, the system may correspond to a robotic device comprising at least one tactile interface and at least one actuated component. The robotic device may generate at least one output indicative of the at least one determined interpretation and provide the at least one generated output as at leastone sensory feedback, robotic gesture, robotic sound, robotic visual output, or any combination thereof.

[0021] wherein at least one output is delivered with at least one controlled temporal pattern including at least one micro-delay or rhythmic pacing variation. The timing-based modulation is technical in nature and does not include diagnosis or treatment of emotional states.

[0022] In another aspect of the present disclosure, an Al-based method for enabling user self-awareness is disclosed. The method includes receiving the at least one type of communication, determining the at least one communication timing pattern without semantic analysis, determining at least one representational mapping and at least one somatic marker based on the at least one determined communication timing pattern, generating at least one response indicative of the at least one representational mapping and the at least one somatic marker, and providing the at least one generated response in real-time. Unlike conventional systems, the present subject matter does not classify emotions using semantic labels but instead mirrors raw multimodal timing signals (tone, rhythm, silence, pacing, and micro-timing variations) within a predetermined temporal window.BRIEF DESCRIPTION OF DRAWINGS

[0023] FIG. l is a schematic block diagram of a system for interpreting multimodal communication patterns based on timing and rhythmic features, in accordance with various embodiments of the present disclosure;

[0024] FIG. 2 illustrates an example of a timing-based response generated by the system of FIG. 1 for a text-based communication, in accordance with an embodiment of the present disclosure;

[0025] FIG. 3 is another example illustration of a timing-modulated response generated by the system of FIG. 1 for the text-based communication, in accordance with an embodiment of the present disclosure;

[0026] FIG. 4 is another example illustration of a response generated by the system of FIG. 1 for an audio-based communication, demonstrating timing and rhythm interpretation, in accordance with an embodiment of the present disclosure;

[0027] FIG. 5 illustrates a schematic block diagram of a method for determining multimodal timing patterns and generating a corresponding response, enabling user self-awareness, in accordance with the embodiment of the present disclosure; and

[0028] FIG. 6A to FIG. 6E illustrate the example categories for categorizing a user by the system of FIG. 1, in accordance with the embodiment of the present disclosure.DETAILED DESCRIPTION

[0029] Reference will now be made in detail to embodiments of the disclosure, examples of which are illustrated in the accompanying drawings. Wherever possible, the same reference numbers will be used throughout the drawings to refer to the same or like parts.

[0030] Referring to FIG. 1, a schematic block diagram of a system 100 for interpreting at least one type of communication is disclosed. Examples of the system 100 may include, but are not limited to, computers, laptops, mobile devices, handheld devices, personal digital assistants (PDAs), tablet personal computers, digital notebook, wearable devices, Al-enabled devices, robotic devices, embedded systems, servers, or any similar electronic device configured to process communication signals.

[0031] In an embodiment, the system 100 includes a bus 105 or other communication mechanism for communicating information, and a processor 110 coupled with the bus 105 for processing information. The system 100 also includes a memory 115, such as a random-access memory (RAM) or other dynamic storage device, coupled to the bus 105 for storing information and instructions to be executed by the processor 110. The memory 115 may store temporary variables or intermediate information generated during execution of the instructions. The system 100 further includes a read-only memory (ROM) 120 or other static storage device coupled to the bus 105 for storing static information and instructions for the processor 110. A storage unit 125, such as a magnetic disk or optical disk, is provided and coupled to the bus 105.

[0032] In an embodiment, the storage unit 125 may store data retrieved from different data sources and one or more machine-learning models. The machinelearning models may correspond to mathematical or statistical models generated from computer algorithms based on timing-based features derived from multimodal communication signals. The data retrieved may include, but is not limited to, text signals, audio signals, image signals, and / or video signals, from which timing features (e.g., pauses, pacing, rhythm, latency, micro-timing variations) may be extracted for real-time evaluation by the processor 110.

[0033] The system 100 can be coupled via the bus 105 to an output unit 130. The output unit 130 may include a display unit, an audio unit, a haptic output unit or any combination thereof The display unit may for example, a liquid crystal display (LCD), a Light Emitting Diode (LED) display, an organic LED (OLED) display, or any other visual display device display for displaying information to the user. Further, the audio unit may be a speaker, such as single channel speaker or multichannel speaker for providing an audio output to the user. The processor 110 is configured to generate a visual output for displaying information of the display unit or audio output for producing audio signals from the audio unit. An input device 135, including alphanumeric and other keys, is coupled to bus 105 for communicating information and command selections to the processor 110. Another type of user input device is a cursor control 140, such as a mouse, a trackball, a trackpad, a touch pad, or cursor direction keys for communicating direction information and command selections to the processor 110 and for controlling cursor movement on the display unit. The input devices 135 or 140 can also be included in the display, for example a touch screen and may also include compatible devices with the display unit such as a stylus. Further, the input devices 135 may include microphones, cameras, or other audio-visual devices capable of receiving multimodal signals.

[0034] Various embodiments are related to the use of system 100 for implementing the techniques described herein. In one embodiment, the techniques are performed by the system 100 in response to the processor 110 executing instructions included in the memory 115. Such instructions can be read into the memory 115 fromanother machine-readable medium, such as the storage unit 125. Execution of the instructions included in the memory 115 causes the processor 210 to perform the process steps described herein.

[0035] The term “machine-readable medium” as used herein refers to any medium that participates in providing data that causes a machine to operate in a specific fashion. In an embodiment implemented using the system 100, various machine-readable media are involved, for example, in providing instructions to the processor 110 for execution. The machine-readable medium can be a storage media. Storage media includes both non-volatile media and volatile media. Nonvolatile media includes, for example, optical or magnetic disks, such as storage unit 125. Volatile media includes dynamic memory, such as the memory 115. All such media are tangible to enable the instructions carried by the media to be detected by a physical mechanism that reads the instructions into a machine.

[0036] Common forms of machine-readable medium include, for example, a hard disk, magnetic tape, optical media such as a CD-ROM or DVD-ROM, a solid-state drive (SSD), flash memory, random-access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), nonvolatile memory (NVM), or any other physical memory device capable of storing instructions.

[0037] In another embodiment, the machine-readable medium can be a transmission medium including coaxial cables, copper wire and fiber optics, wireless communication links, radio frequency signals or optical communication signals including the wires that comprise the bus 105. Transmission media can also take the form of acoustic or light waves, such as those generated during radio-wave and infra-red data communications. Examples of machine-readable medium may include but are not limited to, network-based software distribution, downloadable instruction files, cloud-hosted installation packages, or remote update links. For example, the instructions may be transmitted over a network connection from a remote server to the system 100, received through the communication interface 145, and written into the memory 115 for execution by the processor 110. The bus 105 carries the data to the memory 115, from which the processor 110 retrievesand executes the instructions. The instructions received by the memory 115 can optionally be stored on the storage unit 125 either before or after execution by the processor 110. All such media must be tangible to enable the instructions carried by the media to be detected by a physical mechanism that reads the instructions into a machine.

[0038] In an embodiment, the system 100 may be coupled to a network. The network may include communication networks such as, but not limited to, a Local Area Network (LAN), a Wireless Local Area Network (WLAN), a Wide Area Network (WAN), a cellular network, a mobile data network, internet, a Small Area Network (SAN), and the Internet. The system 100 includes a communication interface 145 coupled to the bus 105. The communication interface 145 provides a two-way data communication coupling to the network. For example, the communication interface 145 may include a network interface card (NIC), Ethernet interface, wireless transceiver, or cellular modem capable of providing a data communication connection to a compatible network. As another example, the communication interface 145 can be a local area network (LAN) card to provide a data communication connection to a compatible LAN. Wireless links can also be implemented. In any such implementation, the communication interface 145 sends and receives electrical, electromagnetic, or optical signals that carry digital data streams representing various types of information.

[0039] In some embodiments, the processor 110 is configured to receive at least one type of communication and determine at least one communication pattern associated with the at least one received type of communication without semantic or contextual analysis of the at least one received type of communication.

[0040] In an embodiment, the at least one type of communication received by the processor 110 may correspond to a textual communication, a video communication, an audio communication, a haptic communication, a virtual reality-based communication, an augmented reality-based communication, a sensor-derived physiological communication signal, or any combination thereof.

[0041] In an embodiment, the at least one communication pattern may include, but not limited to a communication tone, communication rhythm, communicationspeed, at least one communication discontinuity, at least one physiological pattern, communication style, a communication method, a communication pause, a communication inconsistency, a physical manner of communication, at least one timing-based content characteristic, at least one micro-timing variation, at least one latency -based timing feature, or any combination thereof.

[0042] The communicate tone may correspond to a timing-based or prosodic characteristic of the received communication, such as amplitude variation, pitch contour, timing fluctuation, or signal intensity modulation. The communication tone is determined based on measurable timing features without semantic, contextual, or emotional interpretation.

[0043] The communication rhythm may correspond to a temporal pattern associated with the received communication, including timing intervals, pacing regularity, cadence, repetition rate, sequential timing of segments, or other measurable temporal structures. The communication rhythm is determined based on timing information extracted from the received signal and does not rely on linguistic stress patterns, semantic structure, or emotional interpretation.

[0044] The communication speed may be the speed of typing characters or speech rate. The communication discontinuity may be a break, disruption, a pause, or delay in the timing of the received communication signal.

[0045] The physiological pattern may correspond to one or more timing-based physiological signals such as touch timing, breathing timing, pressure variation, or sensor-derived biological timing cues. The physiological pattern is determined from measurable timing features derived from sensors, without interpreting affective meaning, emotional state, or psychological attributes.

[0046] The communication style may correspond to timing-based characteristics of the received communication, such as variability of response intervals, timing regularity, pacing patterns, or distribution of pauses. The communication style is determined from measurable timing features without categorizing behavioral intent or emotional traits.

[0047] The communication method may be textual, verbal, visual, or sign based communication. The communication pause may correspond to a measurable silentinterval or lack of signal in the communication stream. The communication inconsistency may correspond to irregular timing patterns, abrupt timing deviations, or timing variations that differ from previously observed timing patterns in the received communication signal.

[0048] The physical manner of communication may correspond to timing-based characteristics extracted from physical or sensor-derived signals, such as timing of gestures, timing of body-movement events, timing of facial-movement patterns, or timing of touch-based interactions. These characteristics are determined based solely on measurable timing information without semantic or emotional interpretation.

[0049] Further, the communication timing-based content characteristic may include a frequency, a combination, a variation, a timing or a timespan, or any combination thereof, associated with a usage of at least one word, character, action, or any combination thereof, in the at least one type of communication received.

[0050] Thereafter, the processor 110 performs a non-semantic or non-contextual analysis of the at least one determined communication pattern and is configured to determine at least one metaphorical representation corresponding to the determined communication tone, the determined communication rhythm, the determined communication speed, the at least one determined communication discontinuity, the at least one determined physiological pattern, or any combination thereof based on the non-semantic or non-contextual analysis. The at least one metaphorical representation may include at least one metaphorical symbol, template, timingbased symbolic structure, or any non-semantic representational element, any combination thereof. The processor 110 may map the at least one determined communication pattern with a plurality of predefined communication patterns, wherein the plurality of predefined communication patterns is associated with a plurality of predefined metaphorical representations, and the at least one metaphorical representation of the at least one determined communication pattern is determined based on the mapping.

[0051] Further, based on the at least one determined metaphorical representation and the at least determined communication pattern, the processor 110 maydetermine at least one interpretation of the at least one received type of communication and perform at least one action based on the at least one determined interpretation within a predetermined time. The at least one action may include generating an output signal that reflects the timing-based characteristics identified from the received communication.

[0052] The at least one interpretation corresponds to a timing-based characterization of the received communication, derived from measurable temporal features without semantic, emotional, or cognitive interpretation. The interpretation may indicate timing variability, timing regularity, latency patterns, temporal clustering, or other timing-based attributes associated with the communication. In some embodiments, the processor 110 may determine one or more timing-based physiological indicators derived from sensor inputs associated with the received communication.

[0053] Further, the processor 110 may determine a communication language, a communication location, a communication environment, or any combination thereof based on the at least one received type of communication, and the determined interpretation is adapted to the determined communication language, the determined communication location, the determined communication environment, or any combination thereof.

[0054] The communication language may include one or more languages along with dialects corresponding to the one or more languages and location, i.e. geographical location. The communication environment may include region, for example, if the communication language is English, the communication region may be United Kingdome or Australia or U.S.A

[0055] The processor 110 may perform at least one action corresponding to the at least one received type of communication. In an embodiment, the at least one action may include generating at least one output indicative of or based on the at least one determined interpretation. The generated output may include at least one textual response, at least one audio response, at least one visual indication, at least one haptic feedback, or any combination thereof. Further, the generated output may include a plurality of output contents. The processor 110 provides the plurality ofoutput contents at predefined intervals with respect to each other via the output unit 130 The timing of the output may be modulated based on the timing characteristics identified in the received communication.

[0056] In another embodiment, the at least one action may include modulating the at least one generated output based on the at least one determined interpretation by introducing at least one time delay by a predetermined time period at a beginning, during, or after providing the at least one generated output. The processor 110 provides, via the output unit 130, the at least one generated output with timingbased modulation.

[0057] The processor 110 is configured to assign a communication style, a communication method, a communication tone, a communication rhythm, a communication discontinuity, a communication pause, a communication inconsistency, a physical manner of communication, or any combination thereof corresponding to the at least one generated output or each output content of the plurality of output contents based on the at least one determined metaphorical representation, the at least one determined interpretation, or a combination thereof as timing-based output characteristics.

[0058] In an embodiment, the processor receives, via the input device 135 or 140, at least one user information associated with the at least one user and generates at least one output based on the at least one received user information and the at least one determined interpretation, where the user information comprises metadata and non-semantic attributes.

[0059] In an embodiment, the action may be based on a timing-based somatic marker. The processor 110 may determine at least one somatic marker in the at least one received type of communication based on the at least one determined communication pattern, the at least one determined metaphorical representation, or any combination thereof. The processor 110 may generate at least one timing-modulated output corresponding to the identified somatic marker without providing medical, therapeutic, or diagnostic recommendations.

[0060] In an embodiment, when the received communication includes at least one timing-based communication segment, the processor 110 may identify at least onetime delay in the at least one received communication content corresponding to the communication tone, the communication rhythm, the communication discontinuity, the communication pause, the communication inconsistency, or any combination thereof comprise at least one time delay. Thereafter, the processor 110 may assign a weighted priority to the at least one identified time delay corresponding to the communication tone, the communication rhythm, the communication discontinuity, the communication pause, the communication inconsistency, or any combination thereof. The at least one interpretation is determined based on the at least one identified time delay and the assigned weighted priority and the at least one output is generated which is indicative of the at least one determined interpretation or associated with the at least one determined interpretation based on the assigned weighted priority.

[0061] In some embodiments, the system 100 may receive the at least one received type of communication from a plurality of users. The processor 110 identifies each user of the plurality of users based on user-specific timing patterns, signal characteristics, or sensor-derived identifiers, and may generate at least one output common to the plurality of users or unique to each user based on the at least one determined interpretation. The system 100 may include at least one sensor associated with at least one user, and the processor 110 may generate at least one output indicative of the at least one determined interpretation based on at least one input received from the at least one sensor.

[0062] The Al model 150 is trained to determine timing-based communication patterns, determine at least one symbolic representation of the communication patterns, interpret the received communication based on these representations, and perform an action or generate an output accordingly. The Al model 150 operates without semantic or contextual analysis.

[0063] In some embodiments, the Al model 150 may optionally estimate non-semantic user metadata based on timing-derived features of the received communication. Such metadata may include user categories inferred from temporal characteristics of the communication signal, without relying on semantic, emotional, psychological, or physiological interpretation. The Al model 150 mayadapt timing parameters of the generated output based on such non-sensitive metadata where appropriate.

[0064] The generated response is part of the generated output. The response may be textual, audio, visual, or haptic, and may be modulated based on the timing-derived characteristics identified in the received communication.

[0065] In some embodiments, the Al model 150 may be trained using machinelearning methods configured to learn timing-based features extracted from text, audio, video, or sensor-derived signals. The Al model 150 may generate timing-modulated outputs using learned temporal patterns. The Al model 150 does not rely on semantic interpretation, emotional inference, or linguistic meaning when determining communication patterns.

[0066] The Al model 150 may be trained to identify and represent timing-based features such as rhythm, pacing, micro-pauses, silence intervals, and other measurable temporal structures present in communication signals. The learned representations may be used to generate timing-modulated outputs corresponding to the identified temporal characteristics.

[0067] In some embodiments, the Al model 150 may learned using supervised learning methods. Examples of supervised learning may include Linear Regression, Logistic Regression, Support Vector Machines (SVMs), Decision Trees, Random Forests, Neural Network, or any other suitable learning algorithm.

[0068] When the communication is received by the processor 110, the Al model 150 identifies timing-based characteristics present in the communication signal, including rhythm, pace, micro-pauses, silence intervals, or other timing variations. These timing-based patterns may be used to generate output signals without analyzing or inferring semantic content or emotional state.

[0069] Unlike sentiment analysis or emotional classification performed by the conventional systems, the Al model 150 operates solely on timing-based characteristics of the communication signal. The generated output may reflect timing structures such as pacing, rhythm, or silence patterns identified in the received communication, without interpreting emotional meaning, psychological state, or internal conditions of the user.

[0070] The training data may include symbolic or metaphorical representations associated with timing-based communication patterns. These representations may correspond to predefined categories that group together similar temporal features such as pacing patterns, rhythm structures, pause distributions, silence intervals, or other measurable timing characteristics. The training data may also include non-physiological markers derived from timing variations or multimodal timing events.

[0071] Each symbolic representation may correspond to one or more predefined timing-based categories. Examples of categories may include, but not limited to, Fire, Water, Air, Earth, Stone, or any combination thereof. The example categories may be associated with characteristic temporal structures or communication patterns. Based on the detection of one or more symbolic representations, the processor 110 may determine a corresponding interpretation associated with the category, without inferring emotional, physiological, or psychological states.

[0072] Table- 1 illustrates non-exhaustive examples of communication patterns corresponding to different input modalities.Table- 1

[0073] The timing-based categories represent groups of temporal communication patterns and are used to generate output signals with corresponding temporal characteristics. These example categories do not encode emotional labels, physiological states, or psychological interpretations.

[0074] The example categories may represent anchored internal physiological state categories associated with autonomic nervous system regulation of a person. The example categories may function as computational reference states for the system 100 and corresponding to distinct autonomic-related physiological regulation patterns inferred from interaction-based signal features such as temporal rhythm, variability, and stability. The autonomic-related physiological regulation patterns arise within a pre-cognitive timing window, prior to semantic or higher-order cognitive interpretation of the interaction. The example categories do not represent emotional labels, psychological traits, or medical conditions.

[0075] For determining the at least one interpretation, the processor 110 associates the received communication with at least one category selected from the plurality of timing-based categories. Each category is defined by a set of temporal or structural communication features rather than physiological or emotional attributes.

[0076] Each timing-based category may correspond to a predefined set of temporal characteristics such as pacing variability, rhythm patterns, pause distribution, silence intervals, or other structural timing features. These categories function as symbolic groupings of communication timing patterns and do not reflect emotional, psychological, or physiological states.

[0077] For example, FIG. 6A to FIG. 6E illustrate the example categories (Fire, Air, Earth, Water, and Stone) derived from the At-aligned physiological dataset (based on 3,025 sessions of participating users; and ~1.3 million pre-event samples). Each category is represented as a six-dimensional physiological centroid which is based on one or more biometric parameters like temperature (TEMP), ECG, respiration (RESP), Electrodermal Activity (EDA), systolic blood pressure (SBP), and diastolic blood pressure (DBP). The one or more biometric parameters form basis for the system 100 for geometric interpretation.

[0078] FIG. 6A illustrates an example user categorization ‘Fire’ derived by the system 100, when there is increase in the one or more biometric parameters with respect to temperature (TEMP), ECG, and respiration, or there is decrease in the one or more biometric parameters with respect to Electrodermal Activity (EDA),systolic blood pressure (SBP) and diastolic blood pressure (DBP), or any combination thereof, within the time duration of 0.30 to 0.40 seconds. The Fire category may indicate the internal state of the user characterized by heightened autonomic arousal and increased sympathetic nervous system activation, reflected through rapid physiological mobilization, elevated metabolic readiness, and increased internal energy or activation. This state is associated with intensified responsiveness, increased alertness, and accelerated physiological dynamics, without assigning diagnostic, emotional, or medical labels to the user.

[0079] FIG. 6B illustrates an example user categorization ‘Air’ derived by the system 100, when there is decrease in the one or more biometric parameters with respect to temperature and respiration, or there is increase in the biometric parameters with respect to ECG along with minor increase in EDA, or any combination thereof, within the time duration of 0.40 to 0.45 seconds. The Air category may indicate the internal state of the user characterized by heightened cognitive and attentional activity with comparatively lower somatic and metabolic engagement, reflected through increased cardiac activity alongside reduced thermal and respiratory intensity. This state is associated with rapid information processing, elevated mental responsiveness, and increased internal variability, without assigning diagnostic, emotional, or medical labels to the user.

[0080] FIG. 6C illustrates an example user categorization ‘Earth’ derived by the system 100, when there is decrease in the one or more biometric parameters with respect to ECG, or there is increase in the one or more biometric parameters with respect to temperature and respiration with stable rate or any combination thereof, within the time duration of 0.42 to 0.48 seconds. Further, there exists a coherence between temperature, respiration, and ECG. The Earth category may indicate the internal state of the user characterized by physiological stability and autonomic coherence, reflected through synchronized and steady interactions between cardiac, respiratory, and thermal signals. This state is associated with regulated energy expenditure, sustained internal balance, and consistent physiological rhythm, enabling predictable and stable response modulation by the system 100, without assigning diagnostic, emotional, or medical labels to the user.

[0081] FIG. 6D illustrates an example user categorization ‘Water’ derived by the system 100, when there is a steep decrease in the one or more biometric parameters with respect to temperature and respiration, or a steep increase SBP and DBP or any combination thereof, within the time duration of 0.55 to 0.90 seconds. The Water category may indicate the internal state of the user characterized by pronounced autonomic downshifting and increased parasympathetic dominance, reflected through reduced thermal and respiratory activity alongside elevated vascular pressure responses. This state is associated with slowed physiological dynamics, increased internal load or conservation processes, and delayed response timing, the system 100 may adapt the output pacing and intensity, without assigning diagnostic, emotional, or medical labels to the user.

[0082] FIG. 6E illustrates an example user categorization ‘ Stone’ derived by the system 100, when there is a decrease in the one or more biometric parameters with respect to respiration, or a steep increase ECG, EDA, SBP, and DBP, or any combination thereof, for the time duration of more than 1.0 second. The Stone category may indicate the internal state of the user characterized by sustained high autonomic load and prolonged physiological rigidity, reflected through reduced respiratory activity alongside elevated cardiac, electrodermal, and vascular responses over an extended duration. This state is associated with low physiological flexibility, increased internal resistance to change, and delayed system responsiveness, the system 100 may apply conservative, stabilizing, or minimal output modulation, without assigning diagnostic, emotional, or medical labels to the user.

[0083] In some example implementation, the processor 110 may associate the received communication with a category based on detected timing characteristics. For instance, one of the example categories may be selected when the system 100 identifies increased pacing, decreased pause duration, or other measurable timing variations. These associations are based solely on temporal communication features and do not imply emotional or physiological interpretation. The system 100 may generate a response to the received communication from the user based on the categorization of the user.

[0084] The training data may include labeled samples that pair communication signals with corresponding timing-based example categories, timing patterns, symbolic representations, or weighted timing priorities. Labels may describe temporal attributes such as pacing rate, pause intervals, rhythm structure, or silence patterns. The labeled data may also include associated system actions such as timing-modulated textual or audio responses.

[0085] For audio communication, labels may include features such as timing of pitch transitions, durations of silence intervals, rhythm patterns or associated category. For sensor-based communication, labels may include timing of contact events, timing of gesture intervals, or timing-based motion markers. For video communication, the system 100 may utilize timing of movements or time-based changes in visual frames, without interpreting emotional or physiological expressions.

[0086] In an embodiment, a dataset of 5000 or more communication samples annotated with timing features such as pause histograms, syllabic rates, pacing intervals, rhythm density, silence markers, category or any combination thereof may be used to train the Al model 150. The model parameters are optimized to recognize and categorize timing-based communication patterns.

[0087] Based on the timing features detected in the received communication, the trained Al model 150 may generate an output that reflects corresponding temporal characteristics such as pacing, rhythm, or silence intervals. The generated response mirrors timing patterns without interpreting emotional signals, psychological states, or physiological responses.

[0088] The Al model 150 is trained under the principle that communication timing patterns can be identified and represented without reliance on semantic content, emotional labels, or physiological interpretation. The system 100 considers measurable temporal features.

[0089] The output generated by the system 100 may include timing-modulated textual, audio, or visual signals that reflect the temporal characteristics identified in the received communication. These outputs correspond to timing structures suchas rhythm, pacing, or pause intervals and do not convey emotional content or physiological influence.

[0090] In some embodiments, the Al model 150 is configured to identify timingbased communication signals across multiple modalities. These signals may include temporal characteristics such as pacing, rhythm, micro-pauses, silence intervals, and other measurable timing patterns. Based on these timing features, the Al model 150 may determine one or more symbolic representations and associate them with predefined timing-based categories.

[0091] The Al model 150 may generate output signals that reflect the temporal structure of the received communication, such as modulated pacing, rhythm adjustments, or silence intervals. The output reflects timing characteristics rather than semantic content or emotional interpretation.

[0092] In some embodiments, the processor 110 may account for languagespecific timing patterns when generating output. For example, the timing structure of a mirrored phrase or audio signal may vary depending on the linguistic timing conventions of different languages or cultural contexts. The system 100 modulates output timing to maintain coherence with the identified temporal features of the received communication.

[0093] The Al model 150 may be trained using machine learning techniques to recognize and categorize timing-based features extracted from text, audio, video, or sensor-derived signals. The model does not infer emotional, psychological, or physiological states, and does not analyze semantic meaning. Instead, it identifies and mirrors temporal patterns that characterize the communication signal.

[0094] In some embodiments, the system 100 generates responses that reflect timing-based characteristics such as rhythm, pacing, or pause distributions. These timing-modulated outputs are produced without diagnosing, predicting, or influencing emotional or physiological conditions.

[0095] In some embodiments, the system 100 may operate using a predetermined time interval (At) that represents a reproducible temporal window detected across multiple communication modalities. This predetermined time interval corresponds to a measurable delay between (i) the arrival of an external communication signalat the system input and (ii) the system’s generation of an output based on early timing-based features of the received communication pattern.

[0096] The predetermined time interval At may be defined as the interval within which the system 100 identifies characteristic timing variations — such as pacing changes, pause density shifts, micro-timing irregularities, or multimodal timing convergence — prior to performing higher-order semantic or contextual operations. Thus, At enables the system 100 to process early temporal structure before later computational layers engage. The values of the predetermined time At may lie within a range of 70 to 500 milliseconds, or any suitable interval enabling early timing-based detection.

[0097] The predetermined time interval At is based on the temporal onset of emotion in the body of the user, identifying the moment when physiological rhythms begin to lead conscious appraisal. Therefore, the predetermined time interval At may also be referred to as a pre-cognitive timing window . The system 100 may determine the predetermined time interval At by aligning multimodal physiological inputs (for example, tone, HRV, respiration, and micro-movements) on a unified temporal axis to locate this pre-cognitive timing window. The Al model 150 may include a synchronization layer for aligning multimodal physiological inputs within a reproducible window (i.e. At) that precedes semantic or cognitive processing, thereby determining the measurable onset of physiological activation that arises before conscious appraisal. The processor 110 determines At by aligning two or more communication-derived timing patterns on a unified temporal axis and identifying the earliest consistent rhythmic convergence.

[0098] In one implementation, a multimodal reference dataset of physiological recordings collected across a plurality of induced emotional states was analyzed to determine a predetermined time interval At. The recordings were processed using geometric and timing-based measurements to extract raw physiological timing signals across modalities such as respiration, electrodermal activity, cardiac activity, temperature, and blood pressure. From this analysis, modality-specific temporal offsets were identified, corresponding to measurable differences between early autonomic activation and subsequent cognitive appraisal. Autonomicchannels, such as respiration and electrodermal activity, exhibited representative positive At ranges, which form the basis for defining At as a pre-cognitive timing window used by the system 100.

[0099] The physiological recordings displayed natural variability across users, emotional states, and sensor conditions; however, consistent timing patterns were identifiable across modalities. Autonomic channels exhibited detectable temporal leads within the At window, whereas cardiovascular modalities commonly showed At values near zero or within small negative ranges. These modality-specific timing characteristics are incorporated into the system architecture, enabling the processor 110 to align multimodal physiological inputs on a unified temporal axis before any semantic or contextual interpretation occurs. With such alignments the system 100 may determine the onset of a pre-cognitive physiological shift.

[0100] Table 2 provides representative At ranges for several physiological channels. These ranges are illustrative and non-limiting and may vary based on sensor configuration, sampling frequency, environmental factors, or user physiology. The processor 110 or the Al model 150 may reference these timing ranges as part of determining early-state physiological activation or timing-based communication patterns. These ranges represent timing behaviors observed across sets of multimodal physiological recordings and are not tied to any specific dataset size or composition.Table 2 — Representative At Ranges by Physiological Channel (Illustrative, non-limiting, modality-dependent)

[0101] The reference dataset provides an example of how the At interval may be derived and utilized within the system 100. By aligning multimodal physiological signals within the corresponding pre-cognitive timing window, the system 100 can determine when a physiological change begins prior to cognitive appraisal. This enables the generation of timing-based interpretations and outputs that reflect pre-cognitive physiological rhythms. Accordingly, the At interval functions as a system-level timing parameter derived from empirical physiological measurements and used to support early-state detection within the disclosed architecture.

[0102] In some embodiments, At timing offsets may also be considered at the level of recording context rather than individual physiological channels. Recording contexts may include, for example, neutral states, positive engagement states, threat-related states, or high-arousal states. Table 3 below provides representative At ranges for such contexts. These values are illustrative and non-limiting andreflect the general timing behavior that may be observed across multiple physiological modalities within each context. The ranges do not correspond to any fixed dataset and may vary according to user physiology, environmental conditions, and sensor configuration.Table 3 — Representative At Ranges by Recording Context (Illustrative and non-limiting)

[0103] In some embodiments, individual recordings associated with threat-related or anger-related contexts may exhibit variability across modalities. Although the representative mean At ranges for such contexts may include negative values, certain autonomic channels — such as respiration or electrodermal activity — may still show positive At offsets within individual recordings. This indicates that early autonomic timing shifts may occur across a range of emotional or arousal conditions. Accordingly, At may serve as a robust timing parameter that can be applied across different recording contexts for identifying early physiological activation prior to cognitive appraisal.

[0104] In an embodiment, the system 100 may derive physiological timing information from multimodal recordings. Such recordings may include, for example, ECG, electrodermal activity, respiration, systolic and diastolic bloodpressure, and peripheral temperature. Timing features extracted from the physiological modalities may be associated with a plurality of emotional or engagement conditions and used to define representative patterns, i.e. the example categories (such as Fire, Water, Air, Earth, and Stone). The timing relationships obtained from these recordings may further be used to determine the At interval and to associate each category with a corresponding nervous-system activation profile.

[0105] The system 100 may categorize the user based on the received communication and based on presence of one or more communication patterns. The system 100 may determine values of the one or more communication patterns and identify an associated category for the user based on the determined values. Based on the identified category, the system 100 may modulate response parameters such as timing, pacing, intensity, structure, and interaction dynamics to align the response with the user’s detected physiological regulation pattern, without semantic interpretation or diagnostic inference. For each of the example categorizations, the system 100 may have associate a predefined response profile which defines output behavior, enabling adaptive responses while maintaining consistency, predictability, and safety across users and contexts.

[0106] For example, a first predefined response profile corresponding to the Fire category may include reduced response latency, increased response energy, dynamic pacing, and concise yet action-oriented output structure to match elevated physiological activation while avoiding escalation. A second predefined response profile corresponding to the Water category may include increased response latency, slower pacing, softened interaction dynamics, and reduced output intensity to accommodate slowed physiological processing and extended internal regulation timing. A third predefined response profile corresponding to the Air category may include moderate response latency, lighter pacing, flexible or modular output structure, and reduced somatic emphasis to align with heightened cognitive activity and increased internal variability. A fourth predefined response profile corresponding to the Earth category may include steady response timing, balanced pacing, structured and predictable output patterns, and consistent interactiondynamics to reinforce physiological coherence and stability. A fifth predefined response profile corresponding to the Stone category may include extended response latency, minimal output intensity, simplified structure, and conservative interaction dynamics to prevent overload and support gradual physiological stabilization. Thus, the system 100 may modulate the response generated in response to the received communication based on the categorization of the user and corresponding predefined response profile.

[0107] In an embodiment, the Al model 150 may utilize a weighted-sum function to assess the relative contribution of one or more communication patterns in determining a corresponding category (Fire, Water, Air, Earth, or Stone) and, optionally, a modulation state (e.g., spark, steam). The weighted priority may be represented by:c S arSi(t),where Si(t) represents one or more communication patterns (for example, pause duration, pitch slope, silence markers, or timing irregularities), «i represents a weight assigned to each communication pattern, i is the number communication patterns detected by the system 100, and c is a normalization coefficient configured to maintain the resulting score within a bounded numerical range (for example, 0-1).

[0108] Thus, the Al model 150 may determine at least one communication pattern associated with the at least one received type of communication without performing semantic or contextual analysis of the received input. The Al model 150 may assign weights to one or more communication patterns and utilize these weights to compute a weighted priority in real time. The weighted priority may be used by the system 100 to generate an output that mirrors the predicted pre-cognitive timing state of the user’s nervous system.

[0109] The system 100 evaluates the output generated in response to the received communication from the user. The evaluation includes determining a coherence ratio that compares the input communication signal to the mirrored output. The evaluation may be based on the relation:R(t) > 0,where R(t) represents a computed coherence ratio comparing the received communication pattern to the generated mirrored response, and 0 represents a threshold value, which may lie within a predetermined range such as 0.7 to 0.9. When the coherence ratio satisfies the threshold condition, the system 100 may determine that the mirrored response corresponds sufficiently to the received input communication patterns.

[0110] A feedback signal may be generated by the Al model 150. The feedback signal may be represented by AC(x,t). After generating a mirrored response, the Al model 150 may compare the user’s subsequent or ongoing communication signals (such as tone, rhythm, pauses, or optional sensor-derived inputs) with the previously generated response. If the comparison indicates a mismatch or incoherence, the Al model 150 may generate the feedback signal AC(x,t). For example, the Al model 150 may determine a current internal state of the user based on a newly received communication signal and compare the determined state with an expected internal state derived from the previously generated mirrored response. When a mismatch or incoherence is detected, the feedback signal may modify one or more parameters of the next response to be generated. When coherence is detected, the Al model 150 may either refrain from generating a feedback signal or generate a feedback signal indicating coherence.[oni] The Al model 150 may use the feedback signal to adjust a subsequent response in real time. The adjustments performed by the Al model 150 may include modifying micro-delays (St), altering tonal strength, adjusting pacing patterns, or reshaping the timing envelope of generated speech or text. By applying these feedback-based modifications, the Al model 150 may iteratively refine the generated output so that the mirrored response more closely corresponds to the user’s current communication patterns, as reflected in measurable features such as response timing, rhythm, tone contours, micro-pauses, and, when available, optional physiological indicators. The Al model 150 operates as a dynamic adjustment mechanism that reduces incoherent timing deviations and maintains alignment between input and output signals within the system.

[0112] The system 100 includes computational layers inspired by organizational principles of the biological nervous system and configured to align non-semantic input frequencies (including tone, rhythm, pacing, micro-pauses, HRV-related timing features, and micro-movement patterns) on a unified temporal axis. These layers generate coherence-based outputs within the pre-cognitive timing window (predetermined time At), and are distinct from semantic or cognitive inference modules. In some embodiments, the system 100 may derive indicators of physiological activation states from frequency -based communication features without requiring dedicated biosensors, by converting pacing, silence duration, rhythmic structure, and prosodic variations into coded parameters aligned on the temporal axis. Whereas conventional systems rely on sensors such as heart-rate monitors or EEG hardware, the system 100 extracts timing -based physiological cues directly from naturally occurring communication patterns (for example pause ratio, silence markers, prosodic contour, tokens per second, or punctuation bursts) and mirrors such cues without specialized hardware.

[0113] In an illustrative example, the processor 110 receives a text-based communication such as: “I can’t find my backpack! ! ! I looked everywhere and now I’m late.” The Al model 150 may extract communication parameters including, but not limited to: duration of text entry (for example, 6-10 seconds), number of exclamation marks (3), count of characters repeated three or more times (1), average typing or speaking rate (for example, 3.2 words per second), variance in sentence length (high), ellipses count (0), use of uppercase tokens (0), and punctuation ratio (for example, 0.18). The Al model 150 may compute the weighted sum Eoii Si(t), where example weights may include ( / (punctuation bursts)=0.35, a(tokens per second)=0.25, ( / (sentence-length variance)=0.20, with remaining weights distributed among other features. The computed weighted sum may be compared to predefined threshold ranges. If the score falls within a particular range, the Al model 150 may assign the corresponding category Z (for example, Fire, Water, Air, Earth, or Stone). Within each category, sub-ranges may be referenced to determine a modulation sub-state M (for example, spark, steam, or flow). In the present example, the calculated score may fall within the thresholdassociated with the Fire category, and within that category, correspond to a modulation sub-state M = spark.

[0114] The Al model 150 may determine weighted priorities a for one or more communication patterns through empirical tuning and model training, during which feature weights are iteratively adjusted against annotated dialogues until the system 100 produces consistent mirrored outputs. These weights are not fixed constants and may represent illustrative ratios used to prioritize communication patterns for real-time processing.

[0115] In the present subject matter, Z denotes the category (Fire, Water, Air, Earth, Stone) derived from the mappings shown in Tables 1-3. Each category represents a broad classification of the detected communication pattern. M denotes a modulation sub-state within that category. Examples of sub-states may include spark, steam, or flow, which the Al model 150 may use to refine the tone and timing parameters of the mirrored response generated by the system 100. In combination, Z and M provide a two-level mapping structure, where Z indicates the overall category assigned to the received communication pattern, and M specifies parameter adjustments used for response generation.

[0116] For example, the Al model 150 may categorize the detected communication pattern as belonging to the Fire category based on the mappings provided in Table- 1. The Al model 150 may further determine that the modulation profile includes a tone configuration consisting of direct, steady, short clauses, with output timing adjustments implemented as micro-delays (St) within a range such as 60 to 120 milliseconds. An example response generated by the system 100 may include: “You’ve searched and still can’t find it. I’m here. Let’s try one slower breath — where do you feel it in your body right now?” This response may mirror timing-related aspects of the user’s input, including urgency and pacing characteristics, without adding new semantic content. The generated response may include short clauses and micro-delays consistent with the selected modulation sub-state. The timing and tonal adjustments may serve to contrast the elevated pacing in the user’s input, enabling the mirrored signal to maintain alignment with the detected communication pattern.

[0117] When an observed feature is evaluated by the Al model 150, the feature may first be compared against its corresponding T value. The T values act as calibration thresholds derived from statistical analysis of training data, indicating baseline ranges for individual audio features. Each feature may be normalized relative to its r value before being incorporated into the weighted sum Xa.i Si(t). The weights aarepresent a subset of the general weights ai that apply specifically to audio features, where certain audio features such as pitch-slope or micro-silence density may be given higher priority.

[0118] The T value does not modify the weight a itself. Instead, T determines the normalization baseline for each feature before multiplication by its corresponding weight. When an observed feature exceeds its T value, the normalized value increases and contributes more strongly to the weighted sum; when it falls below T, the normalized value decreases and contributes less. Thus, T influences the magnitude of a feature’s input, while a determines its relative weighting across features. In audio-based communication, the late-fusion weights aamay be used to emphasize selected audio characteristics before being integrated with features derived from text or optional physiological inputs.

[0119] The resulting weighted sum Ear Sift) may then be compared against predefined threshold ranges used for category assignment. If the score falls within a given range, the Al model 150 assigns the corresponding category Z (for example, Fire). Within that category, further sub-ranges may be used to determine the modulation sub-state M (for example, steam). The same thresholding framework applies across modalities such as text, audio, and optional sensor features, with weights (a‘, aa, ah, etc.) allowing different modalities to contribute appropriately.

[0120] The T terms (TI, 12, TS, etc.) represent empirically derived threshold values obtained during model training. During training, large sets of annotated audio samples may be analyzed to determine typical feature ranges (for example, pitchslope distribution, amplitude variability, or syllabic timing). These measurements are used to compute T values that function as baseline calibration points for their respective features.

[0121] The present subject matter does not evaluate input channels in isolation. The system 100 may compute a composite communication-pattern signature by synchronizing features derived from multiple modalities on a shared temporal axis. For example, when multiple communication channels are available (such as text timing, typing speed, prosodic pitch, or optional physiological timing indicators), the system 100 may align these temporal features and generate a composite signature that represents the combined timing characteristics of the received communication. The system 100 generates the output which may mirror this composite signature within the predetermined time At. Accordingly, the system 100 operates as a temporal-alignment mechanism rather than a semantic or emotion-classification system.

[0122] In an illustrative example, the Al model 150 may determine that a particular category Z and modulation sub-state M correspond to a tone profile having specific timing characteristics. The timing characteristics may include micro-delays (St) within a predetermined range and adjustments to the attack envelope of an audio output. The attack envelope refers to the rate at which the generated audio signal reaches its peak amplitude. By modifying the attack envelope, the Al model 150 may alter the acoustic onset curve to match the timing properties associated with the assigned category Z and sub-state M. An example response may include a sequence of short clauses and timed pauses consistent with the selected modulation pattern. The output is generated according to the determined tone, pacing, and St parameters without introducing semantic inference.

[0123] In another example, the processor 110 may receive multimodal communication comprising text input, audio input, and optionally physiological timing signals such as heart-rate measurements or heart-rate-variability (HRV) timing data. The inputs may include a text stream aligned with an audio window of a predetermined duration and, where available, a physiological timing window.

[0124] For illustration, the text input may include a typed message with features such as token rate, punctuation frequency, or timing gaps. The accompanying audio segment may include prosodic features such as pitch-slope, amplitude variation, ormicro-silences. Where physiological timing data is provided, HRV inputs may include inter-beat timing intervals measured across a defined duration. These features may be used solely as timing-based indicators and are not interpreted for medical or emotional purposes.

[0125] The system 100 may integrate the inputs by extracting linguistic pacing features from the text stream, prosodic timing features from the audio stream, and optional timing-based indicators from physiological signals. When signals from different modalities diverge, the system 100 may normalize each feature stream against its calibration threshold r before combining them within the weighted sum Sa;- Si(t) . Modality-specific weights (a‘, aa, ah) may determine the relative influence of text, audio, and physiological timing features in the fused representation. The fused score may then be compared to predefined thresholds to assign the category Z and sub-state M.

[0126] Fusion refers to combining features from multiple modalities — text features F‘, audio features Fa, and optional physiological timing features Fh— into a single salience score. Each modality is normalized against its corresponding T value and multiplied by its modality-specific weight (a‘, aa, ah). In an illustrative configuration, weights may be assigned as (0.4, 0.4, 0.2), giving comparable influence to text and audio inputs and lower but non-zero influence to physiological timing data.

[0127] Observed feature values may include, for example: text tokens per second, the number of punctuation marks, audio pitch-slope values, frequency of microsilences, and where provided, HRV timing measures such as RMSSD relative to a baseline. These values may be combined into the fused salience score and compared to the predefined thresholds. Based on this comparison, the category Z (e.g., Fire) and the modulation sub-state M (e.g., spark) may be assigned. The generated response may utilize the corresponding St timing adjustments and tone profile associated with the determined Z and M mappings.

[0128] For multimodal inputs, the Al model 150 may implement a fusion function in which text features (F‘), audio features (Fa), and optional physiological timing features (Fh) are each normalized using their respective T thresholds and combinedaccording to modality-specific weights. This process enables the system 100 to generate a consistent category Z and modulation sub-state M even when the modalities provide differing timing characteristics.

[0129] Values extracted from each modality may include, for example: text pacing, exclamation counts, audio pitch-slope, silence density, and timing-based physiological inputs. These features are combined in the weighted sum Xaj Sj(t), compared to predefined ranges, and mapped to category Z and sub-state M. These mappings define the output timing characteristics, including St ranges and clause structure.

[0130] Based on the category Z assigned using Tables 1-3, the Al model 150 may select an output configuration that corresponds to the mapped timing profile. For instance, category Z = Fire may correspond to a direct tone pattern with St delays in a predetermined range. The Al model 150 may generate text or audio output consistent with the associated modulation sub-state without inferring emotional content or providing semantic interpretation.

[0131] In some embodiments, the system 100 applies a collapse function to normalize feature streams extracted from multiple modalities. The collapse function may convert heterogeneous timing features (such as pause duration, token rate, inter-beat intervals, micro-prosody timing, or facial-motion event rates) into standardized indices defined on a monotonic time base. The underlying raw media (audio waveform, video frames, sensor signals) may be discarded after feature extraction, and only normalized timing indices may be retained for multimodal fusion. The collapse function simplifies integration of multimodal timing data by transforming heterogeneous feature units into common computational representations used by the system 100.

[0132] In some embodiments, the system 100 may be implemented within a robotic device comprising at least one tactile interface and at least one actuated component. The robotic device may incorporate or access the trained Al model 150. The Al model 150 may generate at least one output based on the determined interpretation and provide the output as at least one sensory feedback signal, robotic gesture, robotic sound, robotic visual signal, or any combination thereof.The output may include text, audio, visual indication, or haptic feedback corresponding to the at least one received type of communication. The robotic device may be configured as a toy, humanoid robot, smart speaker, smart display, or any other consumer form factor.

[0133] In some embodiments, the system 100 may be deployed on a remote server and accessed as a chatbot application. Communication between the user and the system 100 may occur via a computing device executing a native application, a web-enabled application, or a browser-based interface. The computing device may transmit communication signals to the system 100 over a network and may receive generated outputs for display or playback. Examples of computing devices include, but are not limited to, mobile phones, laptops, desktop computers, kiosks, and tablets.

[0134] In accordance with the present subject matter, an artificial intelligence (Al) based system 100 for interpreting at least one type of communication is described with respect to FIG. 1. The system 100 comprises the processor 110 and the memory 115 storing instructions that, when executed by the processor 110, cause the processor 110 to receive the at least one type of communication, determine, via the Al model 150, at least one communication pattern associated with the at least one received type of communication, perform a non-semantic or non-contextual analysis of the at least one determined communication pattern, determine, via the Al model 150, at least one metaphorical representation corresponding to the at least one determined communication pattern, the metaphorical representation being mapped to at least one predefined category associated with a predefined neural-response pattern, determine, via the Al model 150, at least one interpretation of the at least one received type of communication based on the at least one determined metaphorical representation and the at least one determined communication pattern, and perform at least one action based on the at least one determined interpretation.

[0135] In an embodiment, the at least one action corresponds to generating, by the processor 110 via the Al model 150, at least one output indicative of or based on the at least one determined interpretation, wherein the at least one output comprisesat least one textual response, at least one audio response, at least one visual indication, at least one haptic feedback, or any combination thereof corresponding to the at least one received type of communication, and providing the at least one generated output. In another embodiment, the at least one action corresponds to modulating, via the Al model 150, the at least one generated output based on the at least one determined interpretation. The processor 110 modulates the at least one generated output by introducing at least one time delay by a predetermined time period at a beginning, during, or after the providing of at least one generated output.

[0136] Further, the processor 110 is configured to identify, via the Al model 150, at least one time delay in the at least one received communication content corresponding to the communication tone, the communication rhythm, the communication discontinuity, the communication pause, the communication inconsistency, or any combination thereof comprise at least one time delay, and assign a weighted priority to the at least one identified time delay corresponding to the communication tone, the communication rhythm, the communication discontinuity, the communication pause, the communication inconsistency, or any combination thereof, and generate, via the Al model 150, at least one output indicative of the at least one determined interpretation or associated with the at least one determined interpretation based on the assigned weighted priority. The processor 110 is configured to determine the at least one interpretation of the at least one received type of communication based on the at least one identified time delay and the assigned weighted priority;

[0137] In an embodiment, the processor is configured to map, via the Al model 150, the at least one determined communication pattern with a plurality of predefined categories of communication patterns, wherein the plurality of predefined categories of communication patterns are associated with a plurality of categories of predefined metaphorical representations or a plurality of categories of predefined neural response patterns, and the at least one metaphorical representation of the at least one determined communication pattern is determined based on the mapping.

[0138] In an embodiment, the processor is configured to determine at least one somatic marker in the at least one received type of communication based on the at least one determined communication pattern, the at least one determined metaphorical representation, a content included in the at least one type of communication, or any combination thereof, and recommend, via the Al model, at least one physiological action corresponding to the at least one determined metaphorical representation, the at least determined interpretation, or a combination thereof.

[0139] In an embodiment, the at least one metaphorical representation includes at least one metaphorical symbol, word, emotion, sound, image, or any combination thereof.

[0140] In an embodiment, the at least one generated output includes a plurality of output contents and the processor is configured to provide, via the at least one output unit, the plurality of output contents at predefined intervals with respect to each other.

[0141] In an embodiment, the processor is configured to assign a communication style, a communication method, a communication tone, a communication rhythm, a communication discontinuity, a communication pause, a communication inconsistency, a physical manner of communication, or any combination thereof corresponding to the at least one generated output or each output content of the plurality of output contents based on the at least one determined metaphorical representation, the at least one determined interpretation, or a combination thereof.

[0142] In an embodiment, the processor is configured to determine a communication language, a communication location, a communication environment, or any combination thereof based on the at least one received type of communication, wherein the at least one determined interpretation is adapted to the determined communication language, the determined communication location, the determined communication environment, or any combination thereof.

[0143] In an embodiment, the at least one determined interpretation is indicative of at least one emotional state, at least one sensory experience, at least one interoceptive sensation, at least one subconscious thought or thought pattern, atleast one mental state, at least one social situation, at least one environmental detail, or any combination thereof associated with at least one user.

[0144] In an embodiment, the at least one received type of communication is associated with a plurality of users, and the processor 110 is configured to identify, via the Al model 150, each user of the plurality of users based on the at least one received type communication, the at least one determined communication pattern, the at least one determined interpretation, or any combination thereof, and generate at least one output common to the plurality of users or unique to each user of the plurality of users based on the at least one determined interpretation.

[0145] In an embodiment, the system 100 may include at least one sensor associated with the at least one user and the processor 110 is configured to generate at least one output indicative of the at least one determined interpretation based on at least one input received from the at least one sensor.

[0146] In an embodiment, the at least one received type of communication may be associated with one or more users and the processor 110 is configured to receive at least one user information associated with the one or more users and generate at least one output based on the at least one received user information and the at least one determined interpretation.

[0147] In an embodiment, the at least one received type of communication is associated with the one or more users and the processor 110 is configured to predict, via the Al model 150, an estimated age of the at least one user based on the at least one type of communication received, the at least one determined communication pattern, the at least one determined interpretation, or any combination thereof, and generate at least one output indicative of the at least one determined interpretation based on the predicted age of the one or more user. The age may be determined based on the tone, pitch, modulation or the time duration of the received communication.

[0148] In an embodiment, the at least one action may be performed by the processor 110 by determining, via the Al model, at least one somatic marker associated with the at least one received type of communication, and recommending, via the Al model 150, at least one physiological actioncorresponding to the at least one determined metaphorical representation, the at least determined interpretation, or a combination thereof.

[0149] In an embodiment, the processor 110 is configured to predict, via the Al model 150, an arrival of an emotional state associated with the at least one user based on the at least one determined communication pattern, and generate at least one response based on the prediction such that the generated response modulates, counterbalances, or adjusts to the predicted emotional state.

[0150] In an embodiment, the processor 110 is configured to perform the at least one action within a predetermined time interval, the predetermined time interval corresponding to the pre-cognitive timing window occurring prior to onset of cognitive or semantic processing.

[0151] In an embodiment, the output is generated within a predetermined time (At, which is an interval between receiving the at least one type of communication and onset of cognitive or semantic processing. Further, physiological activation states of the user are determined from frequency -based communication features without use of dedicated biosensors by translating pacing, silence, and rhythm into coded parameters aligned on the temporal axis.

[0152] In an embodiment, the processor 110 determines the predetermined time (At) by aligning the determined two or more communication patterns of the at least one communication pattern on the temporal axis.

[0153] In an embodiment, the system 100 may be a robotic device (for example, a computing system 410 of FIG. 4) comprising at least one tactile interface and at least one actuated component. The robotic device may have a form factor of an artificial human or a toy or an animal. The robotic device 410 may be configured to generate, via the Al model 150, at least one output indicative of or based on the at least one determined interpretation, wherein the at least one output comprises at least one text response, at least one audio response, at least one visual indication, at least one haptic feedback, or any combination thereof corresponding to the at least one received type of communication; and provide the at least one generated output as at least one sensory feedback, robotic gesture, robotic sound, robotic visual output, or any combination thereof.INDUSTRIAL APPLICATION

[0154] Referring to FIG. 2, an illustrative example of a text-based interaction is shown. In this example, the system 100 is implemented on a mobile device 200, which includes the trained Al model 150. The Al model 150 may initiate communication by presenting a text message 205. A user may enter text input 210 via a touch-based input interface. The Al model 150 may extract communication patterns from the received text, including but not limited to typing pace, punctuation frequency, and pause characteristics. For example, input timing of approximately 3.5 tokens per second and repeated punctuation may correspond to a predefined category within the system’s mapping tables (Tables 1-3). These extracted features are processed without semantic interpretation and are mapped to an internal category according to predefined thresholds.

[0155] Based on the assigned category and modulation sub-state, the Al model 150 may generate a text-based output 215. The output may include timing-based structuring elements such as short clauses, adjustable micro-delays (St), or controlled pacing determined by the assigned category Z and modulation sub-state M. The generated output is displayed on the device 200.

[0156] As illustrated in FIG. 2, the system 100 does not rely on semantic meaning of individual words. Instead, the system 100 extracts frequency -based and timingbased communication patterns such as typing pace, punctuation bursts, and silence intervals. These patterns are mapped to predefined categories (e.g., Fire, Water, Air, Earth, Stone) corresponding to timing and pattern signatures. The response generated by the system 100 is produced within the predetermined time At, defined as an interval between reception of the communication and the onset of higher-order language processing. The At timing ensures that the output is generated according to the system’s pre-cognitive timing alignment framework.

[0157] The Al model 150 may also determine at least one somatic marker based solely on feature references in the received communication. For example, a text reference to a body region may be treated as a somatic marker feature. Based on the determined metaphorical representation or interpretation, the Al model 150may generate an associated action recommendation such as initiating a timed pause, silence interval, breathing-pattern prompt, audio cue, tactile feedback, or other predefined system action. These actions are determined exclusively from the communication patterns, timing features, and assigned category; no medical, diagnostic, or therapeutic interpretation is performed.

[0158] A subsequent text input 220 may include additional timing patterns such as increased token rate or reduced pause duration. If optional sensors are present, additional timing features such as inter-breath intervals or HRV-derived timing data may be incorporated. The Al model 150 may classify these features as indicative of a predefined internal mapping category based on the system’s thresholds. The system 100 generates an output 225 whose pacing, St adjustments, and clause segmentation correspond to the assigned category Z and modulation sub-state M. The generated output is displayed on the mobile device 200. The system 100 may use these timing characteristics for pattern alignment only, without inferring emotional state or performing any cognitive or physiological assessment.

[0159] Referring to FIG. 3, another illustrative example of a text-based interaction is shown. A mobile device 300 may include the trained Al model 150 and may interact with an adult user. The Al model 150 may initiate communication by generating a text prompt 305. The user may provide a first textual input 310 (“Whatever.”) followed by a second textual input 315 after a delay interval of approximately 15 seconds. The Al model 150 may extract communication patterns from both inputs, including clipped phrasing, reduced pacing variability, and extended inter-message delay. These extracted patterns may correspond to a predefined category according to the system’s mapping tables (Tables 1-3). Based on these features, the Al model 150 may assign a category Z and modulation substate M.

[0160] The Al model 150 may generate multiple candidate outputs 320A, 320B based on the assigned category Z and sub-state M. The system 100 may apply a ranking function to select the output with the highest relevance score. The ranking criteria may include pattern-matching proximity, timing alignment, and thresholdbased feature correspondence.

[0161] In some embodiments, the system 100 may associate certain communication patterns with predicted somatic markers that serve as internal control labels for output selection. These labels may be used to select timing parameters, micro-delays (St), prosodic shaping, and other modulation controls. A timing controller AT may integrate multiple rhythmic features to adjust output pacing. The acoustic attack envelope of generated audio may also be modified to shape the onset characteristics of audio responses.

[0162] A third textual communication 325 may include features such as flat pacing, low amplitude variation, or reduced rhythmic variability. The Al model 150 may extract these features, normalize them against calibration thresholds T, and assign a category Z and sub-state M. The Al model 150 may then generate an output 330 whose structure, timing, and pacing are determined according to the assigned category and modulation parameters.

[0163] With reference to FIGS. 2 and 3, the system 100 processes input signals by extracting frequency-based and timing-based communication patterns in real time and generating structured outputs based on those patterns. The system 100 mirrors timing and rhythmic features without performing semantic interpretation or emotional labeling.

[0164] FIG. 4 illustrates another example in which the system 100 processes an audio-based communication 400. A user 405 provides an audio input to the computing system 410 that includes the trained Al model 150. The Al model 150 may extract audio features such as pitch contour, amplitude slope, hesitation markers, and silence distribution. Based on the extracted features, the computing system 410 generates an audio response with corresponding timing, pacing, and prosodic shaping. In some embodiments, the Al model 150 may also produce nonverbal outputs, including visual indicators, gesture animations, or symbolic cues.

[0165] FIG. 5 illustrates a schematic block diagram of a method 500. At step 505, the processor 110 receives at least one type of communication from a user. At step 510, the processor 110 determines one or more communication patterns from the received input without semantic analysis. At step 515, the processor 110 determines at least one metaphorical representation and at least one somatic markerbased on the communication patterns. At step 520, the processor 110 generates at least one output corresponding to the determined representations and markers. At step 525, the processor 110 provides the generated output to the output unit 130 in real time.

[0166] The Conventional systems emphasize semantic interpretation or content-driven analysis and typically classify user inputs into predefined emotional labels such as “happy,” “angry,” or “sad.” Such systems generally prioritize lexical or cognitive processing and may not incorporate non-semantic communication cues. In contrast, the present subject matter operates primarily on non-semantic features including tone, rhythm, pacing, silence, and other communication timing characteristics. The disclosed system 100 processes these features as frequencybased and temporal indicators extracted directly from the communication signal, and generates outputs that are aligned to these extracted patterns rather than to semantic categories. By utilizing non-semantic, timing-based communication features, the system 100 may improve responsiveness and interaction quality relative to content-based approaches.

[0167] The present subject matter differs from conventional content-based Al frameworks by introducing a frequency -first processing architecture in which non-semantic communication characteristics guide output generation. The system 100 utilizes calibrated timing models, audio modulation controls, and pacing adjustments that may be culturally or contextually configurable. In some embodiments, the system 100 may be configured to operate without retaining longterm emotional or interpretive data, thereby enabling pattern-based interaction without reliance on stored user profiles.

[0168] In the foregoing description, various embodiments and examples of the present disclosure have been described for purposes of illustration. A person of ordinary skill in the art will recognize that modifications, substitutions, and variations may be made without departing from the scope of the present disclosure as defined in the accompanying claims. Accordingly, the specification and the figures are intended to be illustrative rather than limiting.

Claims

CLAIMSWe claim:

1. An artificial intelligence (Al) based system for interpreting at least one type of communication, comprising:a processor; anda memory storing instructions that, when executed by the processor, cause the processor to:receive, via an input device, the at least one type of communication; determine, via an Al model, at least one communication pattern associated with the at least one received type of communication, the at least one communication pattern comprising a communication tone, a communication rhythm, a communication speed, at least one communication discontinuity, at least one physiological pattern, or any combination thereof;perform a non-semantic or non-contextual analysis of the at least one determined communication pattern;determine, via the Al model, at least one metaphorical representation corresponding to the at least one determined communication pattern, the metaphorical representation being mapped to at least one predefined category associated with a predefined neural -response pattern;determine, via the Al model, at least one interpretation of the at least one received type of communication based on the at least one determined metaphorical representation and the at least one determined communication pattern; andperform at least one action based on the at least one determined interpretation.

2. The system of claim 1, wherein the communication pattern further comprises a communication style, a communication method, a communication pause, a communication inconsistency, a physical manner of communication, at least one communication content characteristic, orany combination thereof associated with the at least one received type of communication.

3. The system of claim 2, wherein the at least one communication content characteristic comprises a frequency, a combination, a variation, a timing or a timespan, or any combination thereof, associated with a usage of at least one word, character, action, or any combination thereof, in the at least one type of communication received.

4. The system of claim 2, wherein the at least one received type of communication comprises at least one communication content, and the processor is configured to:identify, via the Al model, at least one time delay in the at least one received communication content corresponding to the communication tone, the communication rhythm, the communication discontinuity, the communication pause, the communication inconsistency, or any combination thereof;assign, via the Al model, a weighted priority to the at least one identified time delay corresponding to the communication tone, the communication rhythm, the communication discontinuity, the communication pause, the communication inconsistency, or any combination thereof, wherein the processor is configured to determine the at least one interpretation of the at least one received type of communication based on the at least one identified time delay and the assigned weighted priority;generate, via the Al model, at least one output indicative of the at least one determined interpretation or associated with the at least one determined interpretation based on the assigned weighted priority.

5. The system of claim 1, wherein the at least one action corresponds to:generating, via the Al model, at least one output indicative of or based on the at least one determined interpretation, wherein the at least one output comprises at least one textual response, at least one audio response, at least one visual indication, at least one haptic feedback, or any combination thereof corresponding to the at least one received type of communication; andproviding, via an output unit, the at least one generated output.

6. The system of claim 5, wherein the at least one action corresponds to:modulating, via the Al model, the providing of the at least one generated output based on the at least one determined interpretation, wherein the modulating comprises:introducing, via the Al model, at least one time delay by a predetermined time period at a beginning, during, or after the providing of at least one generated output.

7. The system of claim 1, wherein the at least one type of communication received corresponds to a textual communication, a video communication, an audio communication, a haptic communication, a virtual reality -based communication, an augmented reality-based communication, or any combination thereof.

8. The system of claim 1, wherein the processor is configured to:map, via the Al model, the at least one determined communication pattern with a plurality of predefined categories of communication patterns, wherein the plurality of predefined categories of communication patterns are associated with a plurality of categories of predefined metaphorical representations or a plurality of categories of predefined neural response patterns, and the at least one metaphorical representation of the at least one determined communication pattern is determined based on the mapping.

9. The system of claim 1, wherein the processor is configured to: determine at least one somatic marker in the at least one received type of communication based on the at least one determined communication pattern, the at least one determined metaphorical representation, a content included in the at least one type of communication, or any combination thereof; andrecommending, via the Al model, at least one physiological action corresponding to the at least one determined metaphorical representation, the at least determined interpretation, or a combination thereof.

10. The system of claim 1, wherein the at least one metaphorical representation includes at least one metaphorical symbol, word, emotion, sound, image, or any combination thereof.

11. The system of claim 5, wherein the at least one generated output comprises a plurality of output contents and the processor is configured to provide, via the at least one output unit, the plurality of output contents at predefined intervals with respect to each other.

12. The system of claim 8, wherein the processor is configured to assign a communication style, a communication method, a communication tone, a communication rhythm, a communication discontinuity, a communication pause, a communication inconsistency, a physical manner of communication, or any combination thereof corresponding to the at least one generated output or each output content of the plurality of output contents based on the at least one determined metaphorical representation, the at least one determined interpretation, or a combination thereof.

13. The system of claim 1, wherein the processor is configured to:determine a communication language, a communication location, a communication environment, or any combination thereof based on the atleast one received type of communication, wherein the at least one determined interpretation is adapted to the determined communication language, the determined communication location, the determined communication environment, or any combination thereof.

14. The system of claim 1, wherein the at least one determined interpretation is indicative of at least one emotional state, at least one sensory experience, at least one interoceptive sensation, at least one subconscious thought or thought pattern, at least one mental state, at least one social situation, at least one environmental detail, or any combination thereof associated with at least one user.

15. The system of claim 1, wherein the at least one received type of communication is associated with a plurality of users, and the processor is configured to:identify, via the Al model, each user of the plurality of users based on the at least one received type communication, the at least one determined communication pattern, the at least one determined interpretation, or any combination thereof; andgenerate at least one output common to the plurality of users or unique to each user of the plurality of users based on the at least one determined interpretation.

16. The system of claim 1, wherein the at least one received type of communication is associated with at least one user and the system comprises:at least one sensor associated with the at least one user, wherein the processor is configured to generate at least one output indicative of the at least one determined interpretation based on at least one input received from the at least one sensor.

17. The system of claim 11, wherein the at least one received type of communication is associated with at least one user and the processor is configured to:receive, via an input device, at least one user information associated with the at least one user; andgenerate at least one output based on the at least one received user information and the at least one determined interpretation.

18. The system of claim 1, wherein the at least one received type of communication is associated with at least one user and the processor is configured to:predict, via the Al model, an estimated age of the at least one user based on the at least one type of communication received, the at least one determined communication pattern, the at least one determined interpretation, or any combination thereof; andgenerate at least one output indicative of the at least one determined interpretation based on the predicted age of the at least one user.

19. The system of claim 1, wherein the system corresponds to a robotic device comprising at least one tactile interface and at least one actuated component, and wherein the robotic device is configured to:generating, via the Al model, at least one output indicative of or based on the at least one determined interpretation, wherein the at least one output comprises at least one text response, at least one audio response, at least one visual indication, at least one haptic feedback, or any combination thereof corresponding to the at least one received type of communication; andproviding, via an output unit, the at least one generated output as at least one sensory feedback, robotic gesture, robotic sound, robotic visual output, or any combination thereof.

20. The system of claim 1, wherein the at least one action corresponds to: determining, via the Al model, at least one somatic marker associated with the at least one received type of communication; and recommending, via the Al model, at least one physiological action corresponding to the at least one determined metaphorical representation, the at least determined interpretation, or a combination thereof.

21. The system of claim 1, wherein the processor is configured to:predict, via the Al model, an arrival of an emotional state associated with the at least one user based on the at least one determined communication pattern; and generate at least one response based on the prediction such that the generated response modulates, counterbalances, or adjusts to the predicted emotional state.

22. The system of claim 1, wherein the processor is configured to perform the at least one action within a predetermined time interval, the predetermined time interval corresponding to a pre-cognitive timing window occurring prior to onset of cognitive or semantic processing.

23. The system of claim 5, wherein the output is generated within a predetermined time (At), the predetermined time corresponding to an interval between receiving the at least one type of communication and onset of cognitive or semantic processing, and wherein physiological activation states of the user are determined from frequency -based communication features without use of dedicated biosensors by translating pacing, silence, and rhythm into coded parameters aligned on a temporal axis.

24. The system of claim 22, wherein the processor determines the predetermined time (At) by aligning the determined two or more communication patterns of the at least one communication pattern on a temporal axis.

5. An artificial intelligence (Al) based method for enabling user self- awareness, comprising:receiving, by a processor, at least one type of communication from the user;determining, by the processor using an Al model, at least one communication pattern associated with the at least one received type of communication without semantic analysis of the least one received type of communication;determining, via the Al model, at least one metaphorical representation of the at least one determined communication pattern and at least one somatic marker associated with the user based on the at least one determined communication pattern;generating, via the Al model, at least one response corresponding to the at least one received type of communication from the user, wherein the at least one generated response is indicative of the at least one determined metaphorical representation and the at least determined somatic marker to enable the user self-awareness; andproviding, via the at least one output unit, the at least one generated response in real-time.

Images