Bad emotion reminding system

Abstract

The invention belongs to the technical field of mobile application programs, and provides an unhealthy emotion reminding system which comprises an unhealthy emotion reminding device which comprises a voice collection unit, a voice recognition and analysis unit and a reminding unit, the voice collection unit is provided with a power line, and the voice recognition and analysis unit is connected with the voice collection unit through the power line. The reminding unit is connected to one side of the voice recognition and analysis unit, recording and analysis of a user are achieved through the voice collection unit and the voice recognition and analysis unit, and through the reminding unit in the unhealthy emotion reminding device, the reminding unit can remind the user of the unhealthy emotion in the modes of window popup, vibration, sound equipment and the like. And the user is prompted and helped to realize emotion change in time, so that positive countermeasures are taken, the influence on working efficiency or interpersonal relationship due to out-of-control emotion is avoided, emotion conflict and contradictory upgrading are avoided, and the effect of adjusting the emotion state is realized.

Description

Technical Field

[0001] The invention belongs to the technical field of mobile application programs, in particular to a negative emotion reminder system. Background Art

[0002] With the rapid development of economy, the rapid improvement of living standards, medical and health care standards, and the continuous improvement of social health awareness, individuals' attention to physical and mental health has become an inevitable trend. The fierce competition environment has brought new survival pressures, resulting in the accumulation of negative emotions in learning and life. Physiological or psychological cognitive dissonance is the main cause of emotional problems.

[0003] As the carrier of emotions, individuals can only withstand and resolve limited negative emotions. If they cannot be adjusted and channeled in time, they will lose control of their emotions or even collapse, which will not only have a negative impact on their physical and mental health, but may also damage interpersonal relationships and social harmony.

[0004] At present, although there are some mental health-related applications, most of them focus on emotion monitoring and recording, and lack real-time and effective reminder and intervention mechanisms to help users control and adjust themselves in a timely manner when negative emotions arise.

[0005] To this end, those skilled in the art have proposed a negative emotion reminder system to solve the problems raised by the background technology. Summary of the invention

[0006] In order to solve the above technical problems, the present invention provides a negative emotion reminder system to solve the negative emotion reminder and other problems in the prior art.

[0007] A negative emotion reminder system, comprising a negative emotion reminder device including a voice collection unit, a voice recognition and analysis unit, and a reminder unit, wherein the voice collection unit is provided with a power line, the voice recognition and analysis unit is connected to the voice collection unit through the power line, and the reminder unit is connected to one side of the voice recognition and analysis unit;

[0008] The voice collection unit includes a real-time collection module, a high-quality collection module, a low-resource consumption module, and a continuous monitoring module. The real-time collection module is used to collect and judge the user's voice. The high-quality collection module can clearly and accurately capture the user's voice content and reduce the analysis error caused by poor collection quality. The low-resource consumption module reduces the consumption of device resources while collecting voice information. The continuous monitoring module continuously collects and monitors the user's voice to ensure that any language expression that may cause bad emotions can be captured;

[0009] The speech recognition and analysis unit includes a speech recognition unit and an emotion analysis unit. The speech recognition unit is provided with an intonation and speech speed analysis module. In the speech recognition process, the intonation information and speech speed information of the user are captured simultaneously. The ups and downs of the intonation and the speed of speech are also used to express emotions. The speech information is written into text information in real time for subsequent emotion analysis. The vector representation is calculated using a pre-trained word vector model and compared with the vectors of other words to capture the subtle emotional differences between words.

[0010] The reminder unit, when a negative emotion is detected, sends a reminder to the user through voice or vibration.

[0011] Preferably, in the language acquisition module unit, the real-time acquisition calculation method is:

[0012] Bit rate = sampling rate × number of quantization bits × number of channels;

[0013] Bit rate refers to the amount of data transmitted per second;

[0014] The sampling rate refers to the number of samples extracted from the continuous signal and composed into discrete signals per second;

[0015] Through the above, the sound collection and recording is achieved.

[0016] Preferably, the real-time acquisition module includes a heart rate detector, a blood pressure tester and language acquisition, the heart rate detector is used to detect the user's heart rate, the blood pressure tester is used to detect the user's blood pressure, and the language acquisition is used to detect and record the user's voice and language, and convert the language information into electrical signals.

[0017] Preferably, in the speech recognition analysis unit, the calculation expression in the speech recognition analysis unit is:

[0018] P(W∣X)=P(X)P(X∣W)P(W)∝P(X∣W)P(W);

[0019] Where P(X|W) is the acoustic model, which represents the probability that the acoustic feature vector generated when a given text sequence W is X;

[0020] P(W) is a language model that calculates the probability of a text sequence appearing in the corpus;

[0021] P(X) is the probability of the speech to be decoded, and is usually considered to be a constant fixed value in practical applications.

[0022] Preferably, the reminder unit includes a vibrator, a speaker and a display, and the reminder unit uses the display, the speaker and the vibrator to send a pop-up window to remind the user of negative emotions.

[0023] Preferably, the speech recognition unit includes a sentiment dictionary combined with a word vector, wherein the dictionary contains a large number of words and their corresponding sentiment tendencies, and a pre-trained word vector model captures subtle semantic differences between words to improve the accuracy of sentiment analysis.

[0024] Preferably, the word vector calculation expression in the speech recognition unit is:

[0025] J(θ)=-T1∑t=1T∑-m≤j≤m,j≠0P(wt+j∣wt)

[0026] Where T is the number of words in the sentence, m is the window size, wt is the current word, wt+j is the context word, and p(wt+j|wt) is the probability of predicting the context word, wt+j, given the center word, wt.

[0027] Preferably, the word vector includes contextual understanding, which takes into account factors such as the user's previous conversation content, context and personal preferences, and corrects and optimizes the sentiment judgment results to ensure that the final sentiment analysis results are more accurate and reliable.

[0028] Compared with the prior art, the present invention has the following beneficial effects:

[0029] 1. In daily life, users can use the negative emotion reminder device to monitor their emotional state in real time. Employees may have negative emotions due to work pressure, relationships with colleagues or project progress, especially when facing pressure, frustration or emotional fluctuations. The voice collection unit and the voice recognition and analysis unit can be used to record and analyze the user. The reminder unit in the negative emotion reminder device can use pop-up windows, vibrations, sounds and other methods to timely remind and help users realize changes in their emotions, so that they can take active coping measures to avoid affecting work efficiency or interpersonal relationships due to emotional out-of-control, and avoid emotional conflicts and escalation of contradictions, thereby achieving the effect of adjusting the emotional state.

[0030] 2. For people who already have mental health problems or are undergoing psychotherapy, the negative emotion reminder device of the present invention can be used as an auxiliary tool. It can help users monitor emotional changes more frequently, communicate with psychological counselors or therapists in a timely manner, and adjust treatment plans. At the same time, for some professional groups that require high concentration and stable emotions, the real-time collection unit in the negative emotion reminder can record the users. Through real-time monitoring and reminder functions, these professionals can stay calm and focused at critical moments to ensure work safety and efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 The figure is a flow chart of the negative emotion reminder system of the present invention. DETAILED DESCRIPTION

[0032] The following embodiments of the present invention are described in further detail in conjunction with the accompanying drawings and examples. The following examples are used to illustrate the present invention, but are not intended to limit the scope of the present invention.

[0033] As attached Figure 1 As shown:

[0034] Embodiment 1: The present invention provides a negative emotion reminder system, comprising:

[0035] The negative emotion reminder device includes a voice collection unit, a voice recognition and analysis unit, and a reminder unit. The voice collection unit is provided with a power line, the voice recognition and analysis unit is connected to the voice collection unit through the power line, and the reminder unit is connected to one side of the voice recognition and analysis unit.

[0036] The voice collection unit includes a real-time collection module, a high-quality collection module, a low-resource consumption module, and a continuous monitoring module. The real-time collection module is used to collect and judge the user's voice. The high-quality collection module can clearly and accurately capture the user's voice content and reduce the analysis error caused by poor collection quality. The low-resource consumption module reduces the consumption of device resources while collecting voice information. The continuous monitoring module continuously collects and monitors the user's voice to ensure that any language expression that may cause bad emotions can be captured;

[0037] The speech recognition and analysis unit includes a speech recognition unit and an emotion analysis unit. The speech recognition unit is provided with an intonation and speech speed analysis module. In the speech recognition process, the intonation information and speech speed information of the user are captured simultaneously. The ups and downs of the intonation and the speed of speech are also used to express emotions. The speech information is written into text information in real time for subsequent emotion analysis. The vector representation is calculated using a pre-trained word vector model and compared with the vectors of other words to capture the subtle emotional differences between words.

[0038] The reminder unit, when a negative emotion is detected, sends a reminder to the user through voice or vibration.

[0039] Furthermore, in the language acquisition module unit, the real-time acquisition calculation method is:

[0040] Bit rate = sampling rate × number of quantization bits × number of channels;

[0041] Bit rate refers to the amount of data transmitted per second;

[0042] The sampling rate refers to the number of samples extracted from the continuous signal and composed into a discrete signal per second;

[0043] Through the above, the sound collection and recording is achieved.

[0044] Furthermore, the real-time acquisition includes a heart rate detector, a blood pressure tester and language acquisition. The heart rate detector is used to detect the user's heart rate, the blood pressure tester is used to detect the user's blood pressure, and the language acquisition is used to detect and record the user's voice and language, and convert the language information into electrical signals.

[0045] Further, in the speech recognition analysis unit, the calculation expression in the speech recognition analysis unit is:

[0046] P(W∣X)=P(X)P(X∣W)P(W)∝P(X∣W)P(W);

[0047] Where P(X|W) is the acoustic model, which represents the probability that the acoustic feature vector generated when a given text sequence W is X;

[0048] P(W) is a language model that calculates the probability of a text sequence appearing in the corpus;

[0049] P(X) is the probability of the speech to be decoded, and is usually considered to be a constant fixed value in practical applications.

[0050] Furthermore, the reminder unit includes a vibrator, a speaker and a display, and the reminder unit uses the display, the speaker and the vibrator to send a pop-up window to remind the user of negative emotions.

[0051] Furthermore, the speech recognition unit includes a combination of a sentiment dictionary and a word vector, wherein the dictionary contains a large number of words and their corresponding sentiment tendencies, and a pre-trained word vector model captures subtle semantic differences between words to improve the accuracy of sentiment analysis.

[0052] Furthermore, the word vector calculation expression in the speech recognition unit is:

[0053] J(θ)=-T1∑t=1T∑-m≤j≤m,j≠0P(wt+j∣wt)

[0054] Where T is the number of words in the sentence, m is the window size, wt is the current word, wt+j is the context word, and p(wt+j|wt) is the probability of predicting the context word, wt+j, given the center word, wt.

[0055] Furthermore, the word vector includes contextual understanding, which takes into account factors such as the user's previous conversation content, context, and personal preferences, and corrects and optimizes the sentiment judgment results to ensure that the final sentiment analysis results are more accurate and reliable.

[0056] Furthermore, the present invention is designed for application in life and study. When facing negative emotions, it is very important to provide appropriate comfort and support. Give the other party sufficient listening space, do not interrupt, do not judge, let the other party feel heard and understood, which is the first step to relieve emotions. Use short words to express your understanding of the other party's feelings, encourage the other party to continue sharing their feelings or experiences, and when appropriate, gently put forward different views or solutions, but avoid directly denying the other party's feelings, and guide them to do some relaxing activities, such as deep breathing, meditation, walking or listening to light music, all of which help to relieve tension.

[0057] From the above, it can be seen that the user installs and turns on the bad mood reminder device on the mobile device, records the daily routine in the bad mood reminder device according to the personal life situation, and records various physical conditions of the individual to achieve the necessary settings of the bad mood reminder device. In order to ensure the different work and life states when the individual uses it, the specific reminder method is adjusted, such as voice or vibration, sensitivity, etc., so as to achieve the reminder effect for the user. The user can adjust the frequency and intensity of the reminder according to personal preferences and needs to ensure that the reminder will not disturb the user too frequently, so as to achieve the effect of effective reminder at the critical moment. After the bad mood reminder device is started, the voice collection module starts to collect the user's voice in real time. The user does not need to perform additional operations when using the bad mood reminder device. The voice collection module will automatically collect voice information. The real-time acquisition module in the sound collection will record the user's blood pressure and heart rate, and the real-time acquisition module will record the user's voice and language in different states. The voice recognition and analysis module processes the collected voice and extracts voice features, so that the bad mood reminder device can distinguish the different states of the user, and at the same time determine the user's heart rate, blood pressure, tone and voice decibels when the user's emotions are stable, so as to record the user's emotional stability value, thereby achieving the user's emotional collection and analysis effect, and also record the user's bad emotions in a limited way to ensure the adjustment effect between the user's emotional stability and bad emotions, and realize the comparison and analysis of the pre-established bad emotion model. This module uses advanced voice recognition algorithms and sentiment analysis models. It is worth noting that in the voice recognition algorithm voice recognition analysis unit, the calculation expression in the voice recognition analysis unit is:

[0058] P(W∣X)=P(X)P(X∣W)P(W)∝P(X∣W)P(W);

[0059] Where P(X|W) is the acoustic model, which represents the probability that the acoustic feature vector generated when a given text sequence W is X;

[0060] P(W) is a language model that calculates the probability of a text sequence appearing in the corpus;

[0061] P(X) is the probability of the speech to be decoded, which is usually considered to be a fixed value in practical applications. At the same time, the sentiment analysis model includes sentiment dictionary and word vector. The sentiment dictionary integrates a rich sentiment dictionary, which contains a large number of words and their corresponding sentiment tendencies (such as positive, negative, neutral), etc. The expression calculated using the word vector module is:

[0062] J(θ)=-T1∑t=1T∑-m≤j≤m,j≠0P(wt+j∣wt)

[0063] Where T is the number of words in the sentence, m is the window size, wt is the current word, wt+j is the context word, and p(wt+j|wt) is the probability of predicting the context word wt+j given the central word. The pre-trained word vector module is used to capture the subtle semantic differences between words to improve the accuracy of sentiment analysis. By comparing the words in the user's voice transcription text with the entries in the sentiment dictionary and combining the similarity of the word vectors, the user's emotional state can be preliminarily judged. At the same time, the user's intonation and speaking speed are analyzed: In addition to the text content, the ups and downs of the tone and the speed of speech are also important factors in expressing emotions. At the same time, the user's intonation information is captured. For example, a high pitch and a fast speaking speed may indicate anger or tension, while a low pitch and a slow speaking speed may indicate sadness or calmness. By analyzing these non-text features through algorithms, the judgment of emotional state can be further refined, and the emotional changes in the user's voice can be accurately identified. If the speech recognition analysis module determines that the user is in a bad emotional state (such as anger, sadness, anxiety, etc.), the reminder module will immediately trigger the corresponding reminder. The reminder method can be set to voice prompts, such as "Please pay attention to control your emotions" or "Take a deep breath and stay calm", or it can be set to vibration reminders to adapt to different usage environments and user needs, and at the same time provide a warning for users with bad emotions. The data is stored, and the collected voice data and analysis results are saved, so that users can view their emotional records and change curves at any time to better understand their emotional state and improvement direction. At the same time, the user's bad emotions will be recorded so that the user can control his bad emotions. At the same time, when the user has a bad mood and needs comfort and encouragement, the emotional dictionary in the bad emotion reminder device will give the user a voice prompt, or directly comfort the user in language. When the user finds that his mood is low and confused, the emotional dictionary in the bad emotion reminder device will be based on the user's way of adjusting emotions and relief methods, so that the user is in a good emotional state and ensure the user's life. At the same time, when the user is in a depressed state for a long time, the bad emotion reminder device can record the user's data. These data can also help the user communicate with a psychological counselor or therapist and adjust the treatment plan. The user can reflect on and improve his emotional management method by viewing the emotional records and analysis results. In addition, for some professional groups that need to concentrate highly and stabilize emotions, the recording effect of the bad emotion reminder device allows users to relax. Users can also adjust the settings of the reminder module according to their emotional changes to obtain more personalized reminder services.

[0064] Working principle: users install and open the bad mood reminder device on their mobile devices, record their daily routines in the bad mood reminder device according to their personal living conditions, and record various physical conditions of their bodies to achieve the necessary settings of the bad mood reminder device. In order to ensure different work and life conditions during personal use, the specific reminder methods are adjusted, such as voice or vibration, sensitivity, etc., so as to achieve the reminder effect for the user. Users can adjust the frequency and intensity of reminders according to their personal preferences and needs to ensure that the reminders will not disturb the user too frequently, so as to achieve the effect of effective reminders at critical moments. After the bad mood reminder device is started, the voice collection module begins to collect the user's voice in real time. During the use of the bad mood reminder device, the user does not need to perform additional operations. The voice collection module will automatically collect voice information and the voice The real-time acquisition module in the sound collection will record the user's blood pressure and heart rate, and the real-time acquisition module will record the user's voice and language in different states. The voice recognition and analysis module processes the collected voice and extracts voice features, so that the bad mood reminder device can distinguish the different states of the user, and at the same time determine the user's heart rate, blood pressure, tone and voice decibels when the user's emotions are stable, so as to record the user's emotional stability value, thereby achieving the user's emotional collection and analysis effect, and also record the user's bad emotions in a limited way to ensure the adjustment effect between the user's emotional stability and bad emotions, and realize the comparison and analysis of the pre-established bad emotion model. This module uses advanced voice recognition algorithms and sentiment analysis models. It is worth noting that in the voice recognition algorithm voice recognition analysis unit, the calculation expression in the voice recognition analysis unit is:

[0065] P(W∣X)=P(X)P(X∣W)P(W)∝P(X∣W)P(W);

[0066] Where P(X|W) is the acoustic model, which represents the probability that the acoustic feature vector generated when a given text sequence W is X;

[0067] P(W) is a language model that calculates the probability of a text sequence appearing in the corpus;

[0068] P(X) is the probability of the speech to be decoded, which is usually considered to be a fixed value in practical applications. At the same time, the sentiment analysis model includes sentiment dictionary and word vector. The sentiment dictionary integrates a rich sentiment dictionary, which contains a large number of words and their corresponding sentiment tendencies (such as positive, negative, neutral), etc. The expression calculated using the word vector module is:

[0069] J(θ)=-T1∑t=1T∑-m≤j≤m,j≠0P(wt+j∣wt)

[0070] Where T is the number of words in the sentence, m is the window size, wt is the current word, wt+j is the context word, and p(wt+j|wt) is the probability of predicting the context word wt+j given the central word. The pre-trained word vector module is used to capture the subtle semantic differences between words to improve the accuracy of sentiment analysis. By comparing the words in the user's voice transcription text with the entries in the sentiment dictionary and combining the similarity of the word vectors, the user's emotional state can be preliminarily judged. At the same time, the user's intonation and speaking speed are analyzed: In addition to the text content, the ups and downs of the tone and the speed of speech are also important factors in expressing emotions. At the same time, the user's intonation information is captured. For example, a high pitch and a fast speaking speed may indicate anger or tension, while a low pitch and a slow speaking speed may indicate sadness or calmness. By analyzing these non-text features through algorithms, the judgment of emotional state can be further refined, and the emotional changes in the user's voice can be accurately identified. If the speech recognition analysis module determines that the user is in a bad emotional state (such as anger, sadness, anxiety, etc.), the reminder module will immediately trigger the corresponding reminder. The reminder method can be set to voice prompts, such as "Please pay attention to control your emotions" or "Take a deep breath and stay calm", or it can be set to vibration reminders to adapt to different usage environments and user needs, and at the same time provide a warning for users with bad emotions. The data is stored, and the collected voice data and analysis results are saved, so that users can view their emotional records and change curves at any time to better understand their emotional state and improvement direction. At the same time, the user's bad emotions will be recorded so that the user can control his bad emotions. At the same time, when the user has a bad mood and needs comfort and encouragement, the emotional dictionary in the bad emotion reminder device will give the user a voice prompt, or directly comfort the user in language. When the user finds that his mood is low and confused, the emotional dictionary in the bad emotion reminder device will be based on the user's way of adjusting emotions and relief methods, so that the user is in a good emotional state and ensure the user's life. At the same time, when the user is in a depressed state for a long time, the bad emotion reminder device can record the user's data. These data can also help the user communicate with a psychological counselor or therapist and adjust the treatment plan. The user can reflect on and improve his emotional management method by viewing the emotional records and analysis results. In addition, for some professional groups that need to concentrate highly and stabilize emotions, the recording effect of the bad emotion reminder device allows users to relax. Users can also adjust the settings of the reminder module according to their emotional changes to obtain more personalized reminder services.

[0071] The embodiments of the present invention are provided for the purpose of illustration and description. Although the embodiments of the present invention have been shown and described above, it can be understood that the above embodiments are exemplary and cannot be understood as limitations of the present invention. Ordinary technicians in this field can change, modify, replace and modify the above embodiments within the scope of the present invention.

Claims

1. A negative emotion reminder system, characterized in that: include The negative emotion reminder device includes a voice collection unit, a voice recognition and analysis unit, and a reminder unit. The voice collection unit is provided with a power line, the voice recognition and analysis unit is connected to the voice collection unit through the power line, and the reminder unit is connected to one side of the voice recognition and analysis unit. The voice collection unit includes a real-time collection module, a high-quality collection module, a low-resource consumption module, and a continuous monitoring module. The real-time collection module is used to collect and judge the user's voice. The high-quality collection module can clearly and accurately capture the user's voice content and reduce the analysis error caused by poor collection quality. The low-resource consumption module reduces the consumption of device resources while collecting voice information. The continuous monitoring module continuously collects and monitors the user's voice to ensure that any language expression that may cause bad emotions can be captured; The speech recognition and analysis unit includes a speech recognition unit and an emotion analysis unit. The speech recognition unit is provided with an intonation and speech speed analysis module. In the speech recognition process, the intonation information and speech speed information of the user are captured simultaneously. The ups and downs of the intonation and the speed of speech are also used to express emotions. The speech information is written into text information in real time for subsequent emotion analysis. The vector representation is calculated using a pre-trained word vector model and compared with the vectors of other words to capture the subtle emotional differences between words. The reminder unit, when a negative emotion is detected, sends a reminder to the user through voice or vibration.

2. A negative emotion reminder system as claimed in claim 1, characterized in that: In the language acquisition unit, the real-time acquisition module is calculated as follows: Bit rate = sampling rate × number of quantization bits × number of channels; Bit rate refers to the amount of data transmitted per second; The sampling rate refers to the number of samples extracted from the continuous signal and composed into discrete signals per second; Through the above, the sound collection and recording is achieved.

3. A negative emotion reminder system as claimed in claim 1, characterized in that: The real-time acquisition includes a heart rate detector, a blood pressure tester and language acquisition. The heart rate detector is used to detect the user's heart rate, the blood pressure tester is used to detect the user's blood pressure, and the language acquisition is used to detect and record the user's voice and language, and convert the language information into electrical signals.

4. A negative emotion reminder system as claimed in claim 1, characterized in that: In the speech recognition analysis unit, the calculation expression in the speech recognition analysis unit is: P(W∣X)=P(X)P(X∣W)P(W)∝P(X∣W)P(W); Where P(X|W) is the acoustic model, which represents the probability that the acoustic feature vector generated when a given text sequence W is X; P(W) is a language model that calculates the probability of a text sequence appearing in the corpus; P(X) is the probability of the speech to be decoded, and is usually considered to be a constant fixed value in practical applications.

5. A negative emotion reminder system as claimed in claim 1, characterized in that: The reminder unit includes a vibrator, a speaker and a display, and the reminder unit uses the display, the speaker and the vibrator to send a pop-up window to remind the user of negative emotions.

6. A negative emotion reminder system as claimed in claim 1, characterized in that: The speech recognition unit includes a combination of a sentiment dictionary and a word vector, wherein the dictionary contains a large number of words and their corresponding sentiment tendencies, and a pre-trained word vector model captures subtle semantic differences between words to improve the accuracy of sentiment analysis.

7. A negative emotion reminder system as claimed in claim 1, characterized in that: The word vector calculation expression in the speech recognition unit is: J(θ)=-T1∑t=1T∑-m≤j≤m,j≠0P(wt+j∣wt) Where T is the number of words in the sentence, m is the window size, wt is the current word, wt+j is the context word, and p(wt+j|wt) is the probability of predicting the context word, wt+j, given the center word, wt.

8. A negative emotion reminder system as claimed in claim 1, characterized in that: The word vector includes contextual understanding, which takes into account factors such as the user's previous conversation content, context, and personal preferences, and corrects and optimizes the sentiment judgment results to ensure that the final sentiment analysis results are more accurate and reliable.

Images