Disclosure of Invention
In order to overcome the problems, the inventor of the invention makes a keen study and designs a system and a method for screening a pushing program based on emotion on a mobile terminal, wherein the system acquires the heart beat interval of a user in real time through a cardiac current measuring device, inputs the acquired information into a trained model, so as to acquire the emotion condition of the user in real time, judges whether the user likes the content type of an operation program according to the relation between the operation program of the mobile terminal and emotion change, marks and stores an analysis result, provides a data basis for subsequently judging whether similar program content types need to be pushed or not, and enables the screening result of the pushing program to be more reasonable and accurate, thereby completing the invention.
Specifically, the present invention aims to provide a system for filtering emotion-based push programs on a mobile terminal, which comprises:
an RRI acquisition module 1 for acquiring cardiac beat intervals of a user;
the emotion recognition module 2 is used for acquiring the emotional condition of the user in real time according to the heart beating interval; and
a program analysis module 3 for analyzing types of program contents that the user likes or dislikes to run in the mobile terminal according to the emotional condition of the user; adding a positive mark to the favorite program content type and adding a negative mark to the disliked program content type;
a storage module 4 for storing each program content type and its corresponding tag run in the mobile terminal over a period of time; and
and the screening module 5 is used for intercepting the push program in the mobile terminal in real time, calling the label with the same program content type from the storage module 4 according to the intercepted push program content type in real time, and further selecting to continue intercepting the push program or releasing interception of the push program.
Wherein, the RRI acquisition module 1 comprises a cardiac current measuring device installed on the back shell or the outer sleeve of the mobile terminal.
Wherein the emotional condition comprises a degree of emotional arousal and a degree of emotional valence.
Wherein, in the program analysis module 3, when the user opens the program, the content type of the program is recorded, the emotion recognition module 2 is continuously called to pay attention to the emotion condition of the user in the running process of the program,
when the emotional arousal value of the user is not lower than the middle level in the use period and the emotional effect value of the user is in an ascending trend, the user is considered to like the program content type;
the emotional impact value of the user may be such that the user does not like the program content type when the usage period is flat or trending down.
Wherein, in the storage module 4, for any program content type, a plurality of corresponding marks can be stored;
preferably, in the filtering module 5, after all the marks corresponding to one program content type are called, according to the difference between the number of the positive marks and the number of the negative marks, the pushing program is selected to be continuously intercepted or to be released from being intercepted.
The invention also provides a method for screening the push program based on the emotion on the mobile terminal, which comprises the following steps:
step a, acquiring the heart beat interval of a user in real time through an RRI acquisition module 1 installed on a mobile terminal;
step b, acquiring the emotional condition of the user in real time according to the heart beating interval through an emotion recognition module 2;
c, analyzing the types of program contents which are liked or disliked by the user according to the emotional conditions of the user through the program analysis module 3; adding a positive mark to the favorite program content type and adding a negative mark to the disliked program content type;
d, storing each program content type and corresponding mark thereof operated in the mobile terminal in a period of time through the storage module 4;
and e, intercepting the push program in the mobile terminal in real time through the screening module 5, calling the mark of the same program content type from the storage module 4 according to the intercepted push program content type in real time, and further selecting to continue intercepting the push program or releasing the interception of the push program.
Wherein the emotion recognition module 2 is obtained by:
step 1, collecting physiological data through a collection device, wherein the physiological data comprises heart beating intervals and converting the physiological data into activity indexes of sympathetic nerves and parasympathetic nerves;
step 2, setting an emotion awakening tag and an emotion valence tag, recording specific emotion arousing degree in the emotion awakening tag, recording specific emotion valence in the emotion valence tag, and combining the comprehensive neural activity index data and the emotion tag into basic data;
step 3, adjusting the format of the basic data to obtain basic data with a uniform format, and judging whether the basic data with the uniform format meets the requirements or not;
step 4, selecting available data from the basic data in the unified format meeting the requirements;
and 5, acquiring an emotion recognition module according to the available data in the step 4.
Wherein each integrated neural activity indicator includes one or more of the following data: the activity index of the sympathetic nerve, the activity index of the parasympathetic nerve, the quotient of the activity index of the sympathetic nerve and the activity index of the parasympathetic nerve, the sum of the activity index of the sympathetic nerve and the activity index of the parasympathetic nerve, and the difference between the activity index of the sympathetic nerve and the activity index of the parasympathetic nerve.
Wherein, the step 3 of judging whether the basic data in the unified format meets the requirement comprises the following substeps:
substep 1, dividing all basic data with uniform format into a learning group and a checking group according to a preset proportion at random;
the substep 2, using the data in the learning group to flush the model, verifying the model one by using each data in the inspection group, and respectively recording the verification result of each data in the inspection group;
substep 3, repeating substep 1 and substep 2, wherein the basic data in the unified format once distributed to the test group is not distributed to the test group any more, and each basic data in the unified format is ensured to verify the model flushed by the data in the learned group in the test group until the verification results corresponding to all the basic data in the unified format are obtained;
and substep 4, calculating the total passing rate of all basic data verification results in the unified format, wherein when the total passing rate is greater than 75%, the basic data in the unified format meets the requirement, otherwise, deleting the basic data in the unified format, and repeating the step 1 and the step 2.
The step 4 of obtaining available data comprises the following substeps:
substep a, repeating substeps 1-3 for a plurality of times, and obtaining a test group consisting of different basic data with uniform formats when substep 1 is repeated each time; enabling the basic data in each uniform format to correspond to a plurality of verification results, and then respectively calculating the average passing rate corresponding to the basic data in each uniform format;
substep b, finding and hiding 1 piece of basic data with the lowest average passing rate and in the unified format, executing substeps 1-4 again by using the remaining basic data in the unified format, observing whether the total passing rate is increased compared with that before hiding the data, if the total passing rate is increased, deleting the hidden basic data in the unified format, and executing substep c; if the total passing rate is not improved, recovering the hidden data, selecting and hiding basic data in a uniform format with the second lowest average passing rate, and repeating the above processes until the total passing rate is improved;
and c, after the total passing rate is improved, repeating the substep a and the substep b on the basis of the residual basic data in the unified format, and after the total passing rate is improved, continuously repeating the substep a and the substep b on the basis of the current residual basic data in the unified format until the total passing rate reaches over 85 percent or the deleted basic data in the unified format reaches 15 percent of the basic data in the unified format, wherein the residual basic data in the unified format is the available data.
Wherein, in step c, when the emotional arousal value of the user is not lower than the middle level in the use period and the emotional effect value of the user is in an increasing trend in the use period, the user is considered to like the program content type;
the emotional value of the user is considered to be disliked by the user when the usage period is flat or tends to decrease.
The invention has the advantages that:
(1) according to the emotion-based push program screening system and method for the mobile terminal, the emotion condition of a user holding the mobile terminal can be known in real time, so that the likes and dislikes of the user on the running program on the mobile terminal are analyzed, and a data basis is provided for later interception of the push program;
(2) according to the emotion-based push program screening system and method on the mobile terminal, provided by the invention, the human body data can be obtained under the condition that a user feels no, the quality of the running program is judged according to the human body data, the push program liked by the user is continuously screened, and the use experience of the user is improved.
Detailed Description
The invention is explained in more detail below with reference to the figures and examples. The features and advantages of the present invention will become more apparent from the description.
The word "exemplary" is used exclusively herein to mean "serving as an example, embodiment, or illustration. Any embodiment described herein as "exemplary" is not necessarily to be construed as preferred or advantageous over other embodiments. While the various aspects of the embodiments are presented in drawings, the drawings are not necessarily drawn to scale unless specifically indicated.
According to the emotion-based push program screening system on the mobile terminal provided by the invention, as shown in fig. 1, the system comprises an RRI acquisition module 1, an emotion recognition module 2, a program analysis module 3, a storage module 4 and a screening module 5. The mobile terminal in the application can be an electronic device such as a mobile phone and a palm computer.
In a preferred embodiment, the RRI acquisition module 1 is configured to acquire cardiac beat intervals of a user;
preferably, the RRI acquisition module 1 comprises a cardiac current measuring device or a photoplethysmograph mounted on the back shell or jacket of the mobile terminal. Preferably, the RRI acquisition module may include any device capable of acquiring the cardiac beat interval, for example, the RRI acquisition module 1 may further include a camera disposed on a mobile phone or a tablet computer, the camera captures a picture of a face of a human body, a hairless part of the face is found as a detection area, that is, a forehead and a cheek are used as the detection area, and changes in brightness of images in the area in consecutive pictures are measured to obtain a consecutive array, where the array is a curve describing the heartbeat activity of the human body, the average brightness is higher and represents diastole, that is, an electrocardiographic trough, and the average brightness is lower and represents systole, that is, an electrocardiographic crest; the time interval between two peaks is the heart beat interval.
When the RRI acquisition module 1 includes a plurality of devices for obtaining the cardiac beat interval, the plurality of devices operate simultaneously and supplement each other, so that the cardiac beat interval data of the user can be output correspondingly at each moment as much as possible. When a plurality of devices output heart beat interval data at the same time, the data output by the device with the highest reliability is taken as the standard; the most frequently used device, i.e., the device that outputs the most heart beat interval data, has the most reliable data for a period of time, e.g., 10 minutes.
In a preferred embodiment, the emotion recognition module 2 is configured to obtain the emotional condition of the user in real time according to the heart beat interval;
preferably, the emotional condition includes an emotional arousal degree and an emotional valence degree, and the emotion recognition module 2 outputs the emotional arousal degree value and the emotional valence degree value of the user in real time according to the cardiac beat interval obtained by the RRI acquisition module.
In a preferred embodiment, the program analysis module 3 is configured to analyze whether the user likes or dislikes the program content type running in the mobile terminal according to the emotional condition of the user, add a positive flag to the liked program content type, and add a negative flag to the disliked program content type;
preferably, in the program analysis module 3, when the user opens a new program, the content type of the program is recorded, and the emotional condition of the user during the running process of the program is continuously concerned,
recording that the user likes the program content type when the emotional arousal value of the user is not lower than the middle level in the use period and the emotional effect value of the user is in an ascending trend in the use period;
the emotional impact value of the user records that the user dislikes the program content type when the usage period is flat or trending down.
Preferably, when a program is used and the running time in the foreground is short, for example, below 5s, the program is not considered, and when a program is used and the running time in the foreground is long, for example, above 2 minutes, the running time of the program is intercepted into a plurality of use periods, for example, one use period every 2 minutes, and for the use time between 5s and 120s, also referred to as one use period, each use period can be correspondingly marked.
In the application, the emotional arousal value is divided into three levels, namely a low level, a medium level and a high level, a numerical value corresponding to each level is related to the number of numerical value gears set in a label, lower 30% gears in all the numerical value gears are classified into the low level, higher 30% gears in all the numerical value gears are classified into the high level, and the middle remaining 40% gears are classified into the high level, for example, when 10 gears exist in the label, the corresponding specific numerical value is 1-10, and when the numerical value corresponding to the emotional condition of a user is 1-3, the user is considered to be at the low level; a value of 4 to 7 corresponds to a medium level; a corresponding value of 8-10 is considered to be high, whereby a corresponding emotional arousal level is provided for each period of use.
In the running process of the program analysis module, after a user finishes using one program, the emotional effectiveness value in the using process is called, and the emotional change trend in each using period is further analyzed.
Specifically, it is preferable that the average emotional effect values in the first half of the usage period are calculated and then the average emotional effect values in the second half of the usage period are calculated, respectively, to compare the change trends; further preferably, when the difference value of the emotional valence values of the two last stages is less than or equal to 10% of the extreme value of the label, the change trend is considered to be flat; thereby providing a corresponding trend of emotional valence changes for each period of use.
Preferably, the type of the program content is determined by the type of the program, and the types of the content of different programs are different, such as a news program, the type of the content can be set by the news type, and can also be set according to information such as content duration, content viewpoint, and the like, such as an online shopping program, and the type of the content can be set by the type of the commodity, and can also be set according to information such as commodity price, commodity delivery time, and the like.
In a preferred embodiment, the storage module 4 is configured to store each program content type running in the mobile terminal for a period of time and its corresponding tag.
Preferably, the period of time is preferably 7-15 days, namely, only the information in the latest period of time is reserved in the storage module, and the old information is covered with the new information in real time, so that the system and the method in the application can be adjusted in time along with the interest change of a user, the storage pressure can be reduced, the cost is reduced, and in addition, the time covering method can also automatically adapt to special situations such as the replacement of a user of the mobile terminal and the like.
In the storage module 4, each program content type corresponds to a positive mark and a quantity and a negative mark and a quantity, and the quantity of the positive marks or the negative marks is adjusted in real time according to a mark instruction given by the program analysis module in the storage process.
In a preferred embodiment, in the screening module 5, after all the tags corresponding to one program content type are called, the number of positive tags and the number of negative tags corresponding to the program content type can be obtained; and selecting to continue intercepting the push program or release intercepting the push program according to the difference value of the number of the positive marks and the negative marks.
Preferably, in step 5, the program content type with the positive mark number greater than the negative mark number is selected, and the difference between the positive mark number and the negative mark number is recorded, that is, the mark difference;
intensively calling a plurality of intercepted pushing programs, corresponding program content types and mark difference values within a short time, such as 3 minutes, wherein the corresponding pushing programs with the mark difference values being negative can be continuously intercepted only by considering the condition that the mark difference values are positive;
sequencing the pushing programs according to the sequence of the marking difference values from large to small, and sequentially releasing the interception of the first 3-5 pushing programs according to the sequence.
The invention also provides a method for screening the push program based on the emotion on the mobile terminal, which comprises the following steps:
step a, collecting the heart beat interval of a user through an RRI collecting module 1;
step b, acquiring the emotional condition of the user in real time according to the heart beating interval through an emotion recognition module 2;
c, analyzing the types of program contents which are liked or disliked by the user according to the emotional conditions of the user through the program analysis module 3; adding a positive mark to the favorite program content type and adding a negative mark to the disliked program content type;
d, storing each program content type and corresponding mark thereof operated in the mobile terminal in a period of time through the storage module 4;
and e, intercepting the push program in the mobile terminal in real time through the screening module 5, calling the mark of the same program content type from the storage module 4 according to the intercepted push program content type in real time, and further selecting to continue intercepting the push program or releasing the interception of the push program.
In a preferred embodiment, the emotion recognition module is obtained by:
step 1, collecting physiological data through a collection device, wherein the physiological data comprises heart beating intervals and converting the physiological data into activity indexes of sympathetic nerves and parasympathetic nerves; the cardiac beat interval is also referred to as the R-R interval;
step 2, setting an emotion awakening tag and an emotion valence tag, selecting a specific emotion arousing degree in the emotion awakening tag, and selecting a specific emotion valence in the emotion valence tag;
the emotion tags comprise an emotion awakening tag and an emotion valence tag, and in a more preferred embodiment, the emotion tags further comprise a special state auxiliary tag;
step 3, adjusting the format of the basic data to obtain basic data with a uniform format, and judging whether the basic data with the uniform format meets the requirements or not;
step 4, selecting available data from the basic data in the unified format meeting the requirements;
and 5, acquiring an emotion recognition module according to the available data in the step 4.
On the basis of obtaining the emotion recognition module, inputting the heart beat interval information collected by the RRI collection module into the emotion recognition module, and obtaining the corresponding emotion excitement degree and emotion valence, namely the emotion condition.
In a preferred embodiment, the collecting device comprises a wearable bracelet and a smart watch, and is preferably a cardiac current measuring device mounted on a back shell or a jacket of the mobile terminal. When the physiological data are collected by the collecting device and the label data are recorded, all the data can be transmitted to the remote server in real time for statistical storage, and a storage chip can be integrated in the collecting device for real-time storage and calculation processing.
In a preferred embodiment, in step 1, two sets of data, namely an activity index of sympathetic nerves and an activity index of parasympathetic nerves, are output according to the collected corresponding conversion of each heart beat interval, so that the scheme in the application has finer time granularity.
In step 1, the two nerves jointly influence the heart beat and the periodic mutual pre-image sound of the nerve activity finally constitutes the heart rate variability.
In a preferred embodiment, the emotional arousal tag is provided with a plurality of values capable of representing emotional arousal degrees, and the corresponding values can be selected according to actual conditions, and preferably, the emotional arousal tag is provided with 5-10 value gears, and the closest value gear is selected according to actual conditions of participants. The emotional arousal label is characterized by emotional arousal degree, the lowest numerical value represents complete calmness, and the larger numerical value represents the more violent emotion.
The emotion valence tag is provided with a plurality of numerical values capable of representing emotion valence, the corresponding numerical values can be selected according to actual conditions, preferably, the emotion valence tag is provided with 2-10 numerical value gears, and the closest numerical value gear is selected according to actual conditions of participants. The emotion valence labels indicate the positive and negative degrees of emotion, the lowest value represents the most negative, and the larger value represents the more positive emotion. The data formats in the two emotion valence tags with the same numerical value gear are uniform, and the data formats in the two emotion awakening tags with the same numerical value gear are uniform.
Preferably, the normalized emotional arousal score is adopted in the emotional arousal tag as an original tag score;
preferably, the emotion titer tag adopts the PANAS standard score as an original tag score, wherein the positive emotion: average 29.7, standard deviation: 7.9; negative emotions: average 14.8, standard deviation 5.4.
Further preferably, in both the emotional arousal tag and the emotional valence tag, 10 parts are divided according to the frequency of data distribution by plus or minus 1.96 standard deviation ranges of the numerical range.
Preferably, in step 2, the emotional tags include an emotional arousal tag and an emotional valence tag, which may be provided separately or simultaneously in the form of coordinates or a chart. The emotion awakening tag is used for recording emotion awakening data, and the emotion valence tag is used for recording emotion valence data.
Preferably, in step 2, the integrated neural activity indicator is related to an activity indicator of sympathetic nerves and an activity indicator of parasympathetic nerves, and each integrated neural activity indicator includes one or more of the following data: an activity index of the sympathetic nerve, an activity index of the parasympathetic nerve, a quotient of the activity index of the sympathetic nerve and the activity index of the parasympathetic nerve, a sum of the activity index of the sympathetic nerve and the activity index of the parasympathetic nerve, a difference between the activity index of the sympathetic nerve and the activity index of the parasympathetic nerve, and the like.
In the application, the collection frequency of the comprehensive neural activity index data is high, and 60-90 or even more groups of the comprehensive neural activity index data can be provided every minute.
The emotional tags are collected relatively infrequently, and may be collected once an hour, or 2-5 times a day. Therefore, each emotion tag data corresponds to a plurality of integrated neural activity index data, and one emotion tag data and the plurality of integrated neural activity index data corresponding to the emotion tag data are combined together to form one basic data. Wherein each emotion tag data comprises emotional arousal data and emotional valence data.
In a preferred embodiment, the value steps in the emotional valence tag and the emotional arousal tag may be the same or different, and a mismatch or data misalignment problem may occur during data statistics, for this reason, in step 3, adjusting the format of the basic data mainly includes adjusting the value and the value step in the emotional tag data; specifically, the number of standard numerical value gears is set, if the number of standard numerical value gears is set to 5 numerical value gears, the numerical value gears in the adjustment basic data are adjusted to 5, the gear numerical value selected from the basic data is adjusted to the gear numerical value under the condition of 5 numerical value gears according to the proportion, and the gear numerical value is rounded up when the gear numerical value cannot be divided.
In a preferred embodiment, the step 3 of determining whether the basic data in the unified format meets the requirement includes the following sub-steps:
substep 1, dividing all basic data with uniform format into two groups randomly according to a preset proportion, namely a learning group and a checking group; preferably, the ratio can be 8-10: 1, and more preferably, the ratio of the number of data in the learning group to the number of data in the testing group is 10: 1;
the substep 2, flushing the model by using the data in the learning group, verifying the model one by using each data in the checking group, and respectively recording the verification result of each data in the checking group, wherein preferably, the verification result comprises a verification pass and a verification fail; the verification is passed by bringing the comprehensive nerve activity index data of the basic data in the same format in the test group into the model, and the obtained emotion label data is consistent with the emotion label data in the basic data, namely, the emotion excitement degree and the emotion valence are consistent; the verification failure refers to that the comprehensive nerve activity index data of the basic data in the test group is brought into the model, and the obtained emotion label data is inconsistent with the emotion label data in the basic data, namely, the emotion stimulating degree and/or the emotion titer are inconsistent;
substep 3, repeating substep 1 and substep 2 for multiple times, wherein the basic data in the unified format which is once distributed into the inspection group is not distributed into the inspection group any more, and ensuring that each basic data in the unified format verifies the model which is washed by the data in the learned group in the inspection group until the verification results corresponding to all the basic data in the unified format are obtained;
substep 4, calculating the total passing rate of the verification results of all the basic data in the uniform format, wherein the total passing rate is the ratio of the verification number of the basic data in the uniform format to the verification number of the basic data in the uniform format; when the total passing rate is not more than 75%, the basic data in the unified format is considered to be not in accordance with the basic requirements, all the basic data are abandoned, the step 1 and the step 2 are repeated, and new basic data are obtained again; and when the result in the sub-step 4, namely the total passing rate is more than 75%, the basic data in the unified format is considered to meet the use requirement, and the next step of processing can be carried out.
In a preferred embodiment, the step 4 of obtaining available data comprises the following sub-steps:
and a substep a, rejecting outlier data aiming at each model-parameter combination by a gradient method, and screening out a model with high ecological utility. Specifically, substeps 1-3 in step 3 are repeated for a plurality of times, and each time substep 1 is repeated, a test group consisting of different basic data in a uniform format is obtained, namely all test groups are different; preferably, the substeps 1-3 are repeated for 8-10 times, so that each basic data in the unified format corresponds to a plurality of verification results, and then the average passing rate corresponding to each basic data in the unified format is respectively calculated; the average passing rate corresponding to the basic data in the unified format is the ratio of the number of passing verification in the verification results corresponding to the basic data in the unified format to the total number of the verification results corresponding to the basic data in the unified format.
Step b, finding and hiding 1 case of basic data with the lowest average passing rate and in the unified format, and hiding one case arbitrarily when the average passing rates of multiple cases of basic data with the unified format are consistent and lowest, wherein the hidden data do not participate in any calculation processing before being recovered; finding and utilizing the residual basic data in the unified format to execute substeps 1-4 again, observing whether the total passing rate is increased compared with that before hiding the data, if the total passing rate is increased, deleting the hidden basic data in the unified format, and executing substep c; if the total passing rate is not improved, recovering the hidden data, and selecting and hiding basic data in the unified format with the second lowest average passing rate, wherein if the average passing rates of a plurality of basic data in the unified format are the same and the lowest, the basic data in the unified format with the lowest hit rate can be selected; repeating the processes until the total passing rate is increased;
a substep c, after the total passing rate is increased, repeating the substep a and the substep b on the basis of the remaining basic data in the unified format, and after the total passing rate is increased, continuously repeating the substep a and the substep b on the basis of the current remaining basic data in the unified format until the total passing rate reaches more than 85%, preferably more than 90%; or until the deleted basic data with the uniform format reaches 15% of the total basic data with the uniform format, the remaining basic data with the uniform format is the available data.
Preferably, the models in the substep 2 include most models with supervised learning, and the washing process of the models includes comprehensive judgment of a plurality of supervised models, and the specific washing process includes, but is not limited to, washing methods using linear regression, support vector machine, gradient descent method, naive bayes classification, decision tree classification, AdaBoost, XGBoost, multilayer neural network, and the like. Preferably, the average value of 2 results which are closer to each other in the results of the 3 models of the multilayer neural network with the 3-4 layers, the C4.5 decision tree and the XGboost is used as the output value of each flushing, namely, the multilayer neural network with the 3-4 layers, the C4.5 decision tree and the XGboost are combined into the most preferable model, namely, the model with high ecological utility. Preferably, in the present application, the neural network selects an RBF neural network.
In the step 5, in the process of obtaining the emotion recognition module, the comprehensive nerve activity index data and the emotion awakening data in each available data are spliced into a data segment which is used as a learning material, and the emotion recognition module is obtained through machine learning;
splicing the comprehensive nerve activity index data and the emotion valence data in each available data into a data segment, using the data segment as a learning material, and obtaining an emotion valence prediction model through machine learning; the prediction model comprises the emotional arousal prediction model and an emotional valence prediction model.
In a preferred embodiment, in step 5, in the learning process of the emotion awakening prediction model and the emotion valence prediction model, a multilayer neural network model, a C4.5 decision tree model and an XGboost model of a 3-4 layer structure are established by using the comprehensive neural activity indexes and the label data at the same time, so as to obtain a multilayer neural network model, a decision tree model and an XGboost calculation module model, a combination of the three models is used as the prediction model, and the output of the prediction model is the average value of two closest output values in the outputs of the three models. For example, for a set of data, three models each give an output of 8, 20, and 7, and the output 7 and the output 8 are close to each other, then the output 7 of the final model is, i.e., the average of 7 and 8, and rounded down.
In a preferred embodiment, in steps 1-5, 1000 participants of each age are tracked for 2 weeks to 2 months to obtain tracking data. The physiological data of the participants come from wearable devices such as smart watches and the like and scanning sensors, and the scoring data come from the daily self-evaluation of the participants; the physiological data is continuously tracked for 24 hours in a mode of acquiring 90 seconds of data every 10 minutes; participants were asked to assess their degree of excitement and emotional valence at least 3 times per day in terms of scoring data for the emotional arousal tag and the emotional valence tag.
In a preferred embodiment, the process of converting the cardiac beat interval RRI into sympathetic and parasympathetic outputs of the composite neural activity indicator is:
using Laguerre function recursion to make the dependent variable be a nearest RRI and make the independent variable be 8 Laguerre recursion decomposition terms X, each decomposition term is composed of an unknown coefficient G, an inferable coefficient phi and an RRI value, and the overall estimation expression is as shown in the following formula (one):
where S represents the upper limit of j, the order of the laguerre polynomial, which determines how many RRIs were used in the past to fit an expression, the more the order, the more accurate the result, preferably 9 are used; j represents the order of the orthogonal laguerre discrete time function; g (j, t) represents a coefficient matrix obtained by combining j-order Laguerre polynomials and RRI interval time in t time range, wherein the coefficient in the coefficient matrix is the coefficient of each included RRI, so that a plurality of RRIs are merged into a recursion Laguerre polynomial, and the last RRI is fitted by the past RRIs to form a recursion relation; f (t) represents the inclusion of the calculated position count for a particular interval in the sequence of interval between adjacent heart beats; n denotes the order of the RRIs traced back forward from this RRINumber; RR
F(t)-nRepresenting any RRI, obtained by laguerre polynomial recursion;
an orthogonal laguerre discrete time function representing the j order, obtained by the following formula (two);
alpha is a constant, and the value of alpha is 0.2;
and calculating the nearest RRI, taking 8 RRIs as the RRIs with the same or more in the reverse direction of time, and substituting the RRIs into the RRI combination to form the RRI ═ sigma (i belongs to 0-2) Xi + ∑ (i belongs to 3-8) Xi. 8 unknown coefficients G are solved by using Kalman autoregression. Substituting sigma (i belongs to 0-2) NiGi and sigma (i belongs to 3-8) NiGi respectively represent sympathetic and parasympathetic output values in the synthetic neural activity index. The matched coefficients N are constants 39, 10, -5, 28, -17, 6, 12, 6, -7, -6, -4 respectively.
When the comprehensive nerve activity indexes are respectively brought into the emotion awakening prediction model and the emotion valence prediction model, the following processing is respectively carried out in the two models:
respectively substituting the comprehensive nerve activity indexes into an emotional arousal prediction model for predicting emotional arousal degree and an emotional valence prediction model for predicting emotional valence; the emotion awakening prediction model receives the comprehensive neural activity index, obtains values output by the multilayer neural network model with the 3-4 layer structure, the C4.5 decision tree model and the XGboost calculation module model respectively, selects 2 relatively close values from the three output values, and calculates the average value of the two values to serve as the output result of the emotion awakening model. The emotion valence prediction model also comprises a multilayer neural network model with a 3-4 layer structure, a C4.5 decision tree model and an XGboost calculation module model, after receiving the comprehensive neural activity index, the emotion valence prediction model obtains values respectively output by the multilayer neural network model with the 3-4 layer structure, the C4.5 decision tree model and the XGboost calculation module model, 2 relatively close values are selected from the three output values, and the average value of the two values is obtained and used as the output result of the emotion valence prediction model.
And finally obtaining the corresponding emotional arousal degree and emotional valence degree, namely the emotional condition.
In a preferred embodiment, in step c, analyzing the types of program contents that the user likes or dislikes running in the mobile terminal according to the emotional condition of the user through the program analysis module 3, adding a positive flag to the type of program contents that the user likes, and adding a negative flag to the type of program contents that the user dislikes;
preferably, in step c, when the user opens a new program, the content type of the program is recorded, and the emotional condition of the user during the running process of the program is continuously concerned,
recording that the user likes the program content type when the emotional arousal value of the user is not lower than the middle level in the use period and the emotional effect value of the user is in an ascending trend in the use period;
the emotional impact value of the user records that the user dislikes the program content type when the usage period is flat or trending down.
Preferably, when the running time of a program in the foreground is short, such as less than 5s, the program is not considered, when the running time of a program in the foreground is long, such as more than 2 minutes, the running time of the program is intercepted into a plurality of use periods, such as one use period every 2 minutes, for the use time between 5s and 120s, also referred to as one use period, each use period can be corresponding to a mark, and for the running program with the length between 120s and 240s, the use period is also classified into one use period.
In the application, the emotional arousal value is divided into three levels, namely a low level, a medium level and a high level, a numerical value corresponding to each level is related to the number of numerical value gears set in a label, lower 30% gears in all the numerical value gears are classified into the low level, higher 30% gears in all the numerical value gears are classified into the high level, and the middle remaining 40% gears are classified into the high level, for example, when 10 gears exist in the label, the corresponding specific numerical value is 1-10, and when the numerical value corresponding to the emotional condition of a user is 1-3, the user is considered to be at the low level; a value of 4 to 7 corresponds to a medium level; a corresponding value of 8-10 is considered to be high, whereby a corresponding emotional arousal level is provided for each period of use.
In step c, after the user finishes each program, the emotional effect value in the using process is called, and the emotional change trend in each using period is further analyzed.
Specifically, it is preferable that the average emotional effect values in the first half of the usage period are calculated and then the average emotional effect values in the second half of the usage period are calculated, respectively, to compare the change trends; further preferably, when the difference value of the emotional valence values of the two last stages is less than or equal to 10% of the extreme value of the label, the change trend is considered to be flat; thereby providing a corresponding trend of emotional valence changes for each period of use.
Preferably, the type of the program content is determined by the type of the program, and the types of the content of different programs are different, such as a news program, the type of the content can be set by the news type, and can also be set according to information such as content duration, content viewpoint, and the like, such as an online shopping program, and the type of the content can be set by the type of the commodity, and can also be set according to information such as commodity price, commodity delivery time, and the like.
In a preferred embodiment, in step d, each program content type running in the mobile terminal over a period of time and its corresponding tag are stored by the memory module 4.
Preferably, the period of time is preferably 7-15 days, namely, only the information in the latest period of time is reserved in the storage module, and the old information is covered with the new information in real time, so that the system and the method in the application can be adjusted in time along with the interest change of a user, the storage pressure can be reduced, the cost is reduced, and in addition, the time covering method can also automatically adapt to special situations such as the replacement of a user of the mobile terminal and the like.
In the storage module 4, each program content type corresponds to a positive mark and a quantity and a negative mark and a quantity, and the quantity of the positive marks or the negative marks is adjusted in real time according to a mark instruction given by the program analysis module in the storage process.
In a preferred embodiment, in step e, after all the tags corresponding to one program content type are called, the number of positive tags and the number of negative tags corresponding to the program content type can be obtained; and selecting to continue intercepting the push program or release intercepting the push program according to the difference value of the number of the positive marks and the negative marks.
Preferably, in step e, the program content type with the positive flag number greater than the negative flag number is selected, and the difference between the positive flag number and the negative flag number is recorded, that is, the flag difference;
intensively calling a plurality of intercepted pushing programs, corresponding program content types and mark difference values within a short time, such as 1-3 minutes, wherein the corresponding pushing programs with the mark difference values being negative can be continuously intercepted only by considering the condition that the mark difference values are positive;
sequencing the pushing programs according to the sequence of the marking difference values from large to small, and sequentially releasing the interception of the first 3-5 pushing programs according to the sequence.
Examples
Selecting an experiencer to hold the mobile terminal for 10 minutes, and in the first 7 minutes, running 3 programs on the mobile terminal, wherein the running contents and the running time are as follows:
content providing method and apparatus
|
Time
|
Times news in news program
|
100s
|
Game program for sports game
|
140s
|
Books shop in shopping program
|
180s |
The method comprises the steps that a cardiac current measuring device is arranged on a back shell of the personal mobile terminal, and an emotion recognition module, a program analysis module, a storage module and a screening module are further arranged in the personal mobile terminal;
wherein, the emotion recognition module obtains the following procedures:
selecting 100 participants, continuously tracking all the participants for 35 days, wearing a smart watch capable of collecting heart beat intervals, acquiring continuous 90-second heart beat interval data every 10 minutes by the smart watch, converting the heart beat interval data into activity indexes of sympathetic nerves and activity indexes of parasympathetic nerves, recording the emotion excitement degree in an emotion awakening label and the emotion valence in an emotion valence label 3 times a day, wherein the label comprises 10 numerical gears, recording the average emotion excitement degree and the emotion valence of the participants in the morning of the day, recording the average emotion excitement degree and the emotion valence of the participants in the afternoon of the day, and recording the average emotion excitement degree and the emotion valence of the participants in the evening of the day.
503060 pieces of RRI data are obtained in total, each piece of RRI data is averaged to obtain 254 values, then the obtained values are converted into activity indexes of sympathetic nerves and activity indexes of parasympathetic nerves, 10000 records containing emotional arousal labels and emotional valence labels are obtained by collecting selected data, one emotional label data and a plurality of comprehensive nerve activity index data corresponding to the emotional label data are combined into one basic data, and 10000 basic data are formed in a conformal mode.
And randomly dividing all 10000 pieces of basic data into 10 parts, wherein one part is used as a test group, the other parts are used as learning groups, flushing the model through the learning groups, verifying the model by using data in the test groups to obtain a verification result of each test group data, using the data in the other parts as the test groups, repeating the steps for 10 times, ensuring that each data is distributed to the test groups once, namely each data obtains a corresponding verification result, solving that the total passing rate is 77 percent and is higher than 75 percent, and carrying out next processing.
And eliminating abnormal data in the basic data to obtain usable data, specifically,
calculating the average passing rate, dividing all basic data into 10 parts again, wherein one part is used as a test group, the other parts are used as learning groups, flushing the model through the learning groups, and verifying the model by using the data in the test groups to obtain the verification result of each data; then, the checking group and the learning group are redistributed, and the process is repeated for at least 100 times, so that each basic data is divided into the checking group for at least 10 times, namely each basic data obtains 10 corresponding verification results, and further the average passing rate of each basic data is obtained;
finding and hiding 1 piece of basic data with the lowest average passing rate, utilizing the remaining 9999 pieces of basic data to execute the process of obtaining the average passing rate and the total passing rate again, observing whether the total passing rate is increased compared with that before hiding the data, and deleting the hidden basic data with the unified format if the total passing rate is increased; if the total passing rate is not improved, recovering the hidden data, selecting and hiding the basic data with the second lowest average passing rate, and repeating the process of obtaining the total passing rate until the total passing rate is improved;
and deleting the hidden data after the hit rate is increased, continuously executing the process of obtaining the average passing rate on the basis of the rest basic data, calculating the average passing rate corresponding to each basic data, searching and hiding the data with the lowest average passing rate, obtaining the total passing rate on the basis of the data with the lowest average passing rate, and continuously repeating the removing process.
And after the hit rate is increased, deleting the hidden data, and continuously repeating the process on the basis of the rest basic data. The remaining data when the total throughput reaches 90% is called usable data.
An emotional arousal prediction model and an emotional valence prediction model are obtained from the available data, and in particular,
using available data to flush an RBF neural network with a 4-layer structure to obtain an RBF neural network model, using available data to flush a C4.5 decision tree to obtain a C4.5 decision tree model, and using available data to flush an XGboost calculation module to obtain an XGboost calculation module model, wherein the three models are combined to form a prediction model; when the prediction model receives a new comprehensive neural activity index, copying the received information into 3 parts, and respectively transmitting the 3 parts to the RBF neural network model, the C4.5 decision tree model and the XGboost calculation module model, wherein the output value of the prediction model is the average value of 2 closer values in the output of the 3 models given by the three models, so that an emotion awakening prediction model and an emotion valence prediction model are obtained, and the prediction models are the emotion recognition modules.
The specific implementation process comprises the following steps:
detecting in real time by the amperometric device the heart beat interval of the experiencer within the first 7 minutes is shown in fig. 2; in the figure, the abscissa represents seconds, and the ordinate represents the interval of heart beats in milliseconds;
inputting the detected heart beat interval data into an emotion recognition module to obtain the emotional arousal degree and the emotional valence of the experiencer in the first 7 minutes, wherein the emotional arousal degree and the emotional valence are shown in the figure 3 and the figure 4 respectively;
analyzing the types of the program contents which the experiencer likes or dislikes to run in the mobile terminal through the program analysis module, and the results are as follows:
and storing the result in a storage module in real time, wherein the information stored in the storage module is as follows:
the time administration news in the news program, the positive mark number is 0, and the negative mark number is 1;
in the game program, the number of positive marks is 0, and the number of negative marks is 1;
the number of book stores in the shopping program is 1 in the affirmative and 0 in the negative.
In the last 3 minutes, the push program in the mobile terminal comprises: the method comprises the following steps that time-administration news in news programs, competitive games in game programs and book stores in shopping programs are intercepted by a screening module firstly, the screening module calls information from a storage module, only the pushing program of the book stores in the shopping programs can obtain a mark difference value, so that the interception of the pushing program of the book stores in the shopping programs is released, and the pushing program is displayed in a mobile terminal: the experiencer can continue to apply the program to the book store in the shopping program.
After completing the trial of the mobile terminal for 10 minutes, the experiencer fills in an evaluation table, whether the pushing program likes in 3 minutes or not is evaluated, and the evaluation result of the experiencer is like.
And further inviting 99 experiences, wherein each experience uses the mobile terminal for 10 minutes, and the storage module of the mobile terminal is emptied before each use.
Of the 100 experiences, 16 were preferred for the first 7 minutes, 22 were preferred for the first 7 minutes,
the remaining 62 experiences liked one or two of the three programs given in the first 7 minutes; and extracting the final evaluation results of the 62 experience persons, wherein the evaluation results are 55 parts which are liked, 5 parts which are disliked and 2 parts which are uncertain, so that the accuracy of the emotion-based push program screening method on the mobile terminal is more than 88.7%.
The present invention has been described above in connection with preferred embodiments, but these embodiments are merely exemplary and merely illustrative. On the basis of the above, the invention can be subjected to various substitutions and modifications, and the substitutions and the modifications are all within the protection scope of the invention.