CN109192310A - A kind of undergraduate psychological behavior unusual fluctuation scheme Design method based on big data - Google Patents
A kind of undergraduate psychological behavior unusual fluctuation scheme Design method based on big data Download PDFInfo
- Publication number
- CN109192310A CN109192310A CN201810824392.1A CN201810824392A CN109192310A CN 109192310 A CN109192310 A CN 109192310A CN 201810824392 A CN201810824392 A CN 201810824392A CN 109192310 A CN109192310 A CN 109192310A
- Authority
- CN
- China
- Prior art keywords
- data
- behavior
- psychology
- information
- unusual fluctuation
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
- G16H50/30—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for calculating health indices; for individual health risk assessment
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H20/00—ICT specially adapted for therapies or health-improving plans, e.g. for handling prescriptions, for steering therapy or for monitoring patient compliance
- G16H20/70—ICT specially adapted for therapies or health-improving plans, e.g. for handling prescriptions, for steering therapy or for monitoring patient compliance relating to mental therapies, e.g. psychological therapy or autogenous training
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
- G16H50/70—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for mining of medical data, e.g. analysing previous cases of other patients
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H80/00—ICT specially adapted for facilitating communication between medical practitioners or patients, e.g. for collaborative diagnosis, therapy or health monitoring
Abstract
The undergraduate psychological behavior unusual fluctuation scheme Design method based on big data that the present invention relates to a kind of, including the multidisciplinary theory such as integrated use systematic science, information science, medicine and psychology, it is merged based on multi-source heterogeneous information, collects virtual-real space Psychology and behavior characterization information；By big data association analysis, the mapping relations model of college student Psychology and behavior Yu behavior characterization information is established；By collecting Intelligent campus big data, appropriate image procossing, text mining, deep learning, incremental learning scheduling algorithm are selected, Psychology and behavior unusual fluctuation monitoring and warning model is established in excavation；Based on Hadoop and Tensorflow big data analysis platform, college student Psychology and behavior unusual fluctuation monitoring and warning software is researched and developed.Compared with existing system, the individual students for detecting Psychology and behavior unusual fluctuation of present invention science have many advantages, such as that accuracy is high, dynamic property is good.
Description
Technical field
The present invention relates to a kind of monitoring and warning and interfering system more particularly to a kind of undergraduate psychological rows based on big data
For unusual fluctuation scheme Design method, belong to big data application field.
Background technique
According to WHO Report, depression is in whole world disease incidence about 11%, it has also become endangers the of human health
Four big diseases are likely to become to the year two thousand twenty and are only second to the cardiopathic second largest disease.In China, incidence of depression is up to 7%,
And because it is found that insufficient with understanding not in time, treatment rate is only 20%.Because committed suicide event caused by depression takes place frequently.Colleges and universities
Student is as specific group active and sensitive in social life, when all experiencing the youth of great variety in physiology, psychology
Phase, Psychological Health Problem are more prominent compared with other crowds.
Currently, the mental health services of college student also rest on " passive " mode, it is main to pass through traditional questionnaire issuer
Formula or student, which seek advice from or go to see a doctor to Mental Health Counseling center, finds psychological unusual fluctuation individual.But by manpower and material resources institute
Limit, psychological study person can not obtain the data of covering research object entirety for a long time, and also inconvenience is to individual mind health shape
State variation carries out follow-up study, it is difficult to carry out timely pro-active intervention to Psychology and behavior unusual fluctuation individual.
With the development of the technologies such as sensor, high speed network, mobile interchange, cloud computing, artificial intelligence, as support
The research in big data technology this ancient and complicated field to psychology is also gradually unfolded.
In patent " a kind of Mental health evaluation system Internet-based " (number: CN201610808709.3), Liu Zhen
It is bright to propose a kind of Mental health evaluation system Internet-based.In system, cloud database is for storing known sample
The factor score of middle psychology test scale；Mental health evaluation model is established using RBF neural network algorithm.RBF neural
After model assesses new individual psychological health states, assessment result is uploaded to cloud.The system is still based on traditional
Psychological test table is as a result, objective evaluation and follow can not be carried out to psychological health states.
In patent " a kind of psychological health states appraisal procedure " (number: CN201210576344.8), the propositions such as Zhu Ting is encouraged
A kind of method carrying out psychological health states assessments using machine learning.The realization step of the appraisal procedure are as follows: firstly, being based on
Individual networks behavioural characteristic and Demographics in known sample, are established and psychology of the training based on network behavior feature is strong
Health status assessment model；Secondly the network behavior feature and Demographics of new individual are obtained；Preferably, according to above-mentioned foundation
Assessment models, obtain the psychological health states of the new individual.Advantage is a cancellation subjective factor and assesses psychological health states
Influence, the disadvantage is that behavioral data source it is single, excavate it is not thorough enough, psychological health states assessment accuracy not can guarantee.
Existing to exert in conclusion although the study on psychological health based on big data technology has obtained some concerns
Description human psychology-behavior theoretical model is set up in power and research not yet, there are data sources single, theory analysis less than
Position, research contents not comprehensively, data mining is not thorough, assesses and only focus on current psychological health states and be not concerned with development trend, comment
Estimate model accuracy and pay attention to the problems such as insufficient, inherently sees or rest on the static evaluation stage of psychological health states.
Find to establish Psychology and behavior unusual fluctuation monitoring and warning model using big data technical method not yet at present to send out in time
Existing Psychology and behavior unusual fluctuation individual and the patent for carrying out pro-active intervention.
Summary of the invention
The purpose of the present invention is overcome defect existing for above-mentioned existing system and provide a kind of data are comprehensive, accuracy is high,
Dynamic property is good, the complete undergraduate psychological behavior unusual fluctuation monitoring and warning of system architecture and interference method and system.
The purpose of the present invention can be achieved through the following technical solutions:
A kind of undergraduate psychological behavior unusual fluctuation scheme Design method based on big data, the system include virtual
Space-entity space Psychology and behavior characterize data fused layer, cloud storage platform, Psychology and behavior-behavior characterize data mapping model,
Psychology and behavior unusual fluctuation monitoring and warning model, Psychology and behavior unusual fluctuation monitoring and warning software, the Virtual Space-entity space psychology
Behavior characterize data fused layer, cloud storage platform, Psychology and behavior-behavior characterize data mapping model, Psychology and behavior unusual fluctuation monitoring
Early-warning Model, Psychology and behavior unusual fluctuation monitoring and warning software are sequentially connected, the Psychology and behavior unusual fluctuation monitoring and warning software and cloud
Storage platform connection；
The Virtual Space-entity space Psychology and behavior characterize data fused layer integrates multiple information sources, utilizes existing number
According to Data Preprocessing Technologies such as cleaning, integrated, transformation, reduction, selection, segmentation and compressions, and suitable for big data scene
Preprocessing Algorithm obtains pretreated behavioural information data and its characteristic attribute；
The multiple information sources include mobile Internet, information platform, image capturing system, Internet of Things etc.；
The data prediction includes the following steps:
Step1: data scrubbing is carried out to the data from multiple information sources, by filling in missing values, smooth noise data, knowledge
Inconsistency that is other or deleting outlier solution data；
Step2: carrying out data integration, and all data acquisition systems from multiple data sources for same individual are got up simultaneously
Take measures to avoid redundancy when data integration, if it is necessary, Step1 can also be carried out again；
Step3: carrying out hough transformation, carries out simplifying expression to data set；
The Preprocessing Algorithm is theoretical with integrated study, proposes a kind of effective data prediction integrated framework；
The behavioural information includes diet, sleep, movement, study, sleep, medical, mood etc.；
The pretreated behavioural information of cloud storage platform distributed storage；
The Psychology and behavior-behavior characterize data mapping model utilizes cloud storage platform by big data correlation analysis
In behavioural information, extract behavior characterize data relevant to Psychology and behavior in students psychology behavioral data, further eliminate special
Correlation between sign reduces the garbage in feature, improves the value density of sample；
The students psychology behavioral data includes student's essential information, student and job information, life information and virtual society
Area's information etc.；
The students psychology behavior includes learning behavior, dietary behavior, motor behavior, internet behavior etc.；
The Psychology and behavior unusual fluctuation monitoring and warning model includes big data analysis frame packet belonging to big data analysis frame
Data Layer, characteristic layer and simulation layer are included, the data Layer, characteristic layer and simulation layer are sequentially connected；
The data Layer cleans all kinds of Psychology and behavior characterize datas, is designed, and proposes filtering type, packaging type, insertion
Formula method selects crucial behavior characterize data character subset, removes unrelated and redundancy feature；
The characteristic layer improves existing a variety of learning models, designs applicable deep learning model, receives multi-source more
The feature of mode carries out the Fusion Features of Psychology and behavior characterize data under big data environment；
The model layer for current machine learning algorithm is more, comparative test time-consuming the problems such as, by data set label generation
For data set as mode input, selection algorithm is exported using the algorithm index of measure algorithm superiority-inferiority as model, reduces modeling
The selection of method and testing time establish Psychology and behavior to carry out decision level Model Fusion with the strategy such as calculation for the national games, integrated study
Unusual fluctuation monitoring and warning model, tracking discovery Psychology and behavior unusual fluctuation individual；
Corresponding algorithm includes text mining, image procossing, prediction modeling, incremental learning etc.；
The Psychology and behavior unusual fluctuation monitoring and warning software be based on Hadoop and Tensorflow big data analysis platform into
Above-mentioned model and algorithm are realized in row research and development, software insertion, specifically:
Step1: campus informatization system data (personnel system, card system, education administration system, school are merged in data Layer
Hospital system, postgraduate's system, library etc.), sensing data (classroom camera data etc.), internet data (online note
Record, social networks behavior etc.) etc.；
Step2: decision-making level is based on above-mentioned model, and dynamically track finds Psychology and behavior unusual fluctuation individual；
Step3: it is oriented in service layer to Psychology and behavior unusual fluctuation individual and recommends psychological consultation article, daily work and rest planning, drink
The information such as food and type of sports push Psychology and behavior unusual fluctuation individual information to students community, counsellor, psychological consultation center etc..
Detailed description of the invention
Fig. 1 is towards college student Psychology and behavior unusual fluctuation monitoring and warning and interfering system frame；
Fig. 2 is Psychology and behavior-behavior characterize data mapping model；
Fig. 3 is Psychology and behavior unusual fluctuation monitoring and warning model；
Fig. 4 is systems technology route；
Fig. 5 machine learning algorithm Selection Framework.
Specific embodiment
The present invention is described in detail with specific embodiment below in conjunction with the accompanying drawings.The present embodiment is with technical solution of the present invention
Premised on implemented, the detailed implementation method and specific operation process are given, but protection scope of the present invention is not limited to
Following embodiments.
In order to make it easy to understand, briefly introducing the principles of science based on the present invention first.
The development and progress of big data technology is changing the every aspect of human society from depth and range, in science
The scientific research of " fourth normal form (4nf "-based on big data after experimental science, induction and conclusion, Computer Simulation is expedited the emergence of in field
Normal form.Under this background, the method that scientists attempt maintenance data science explains that the collapse of the economy, financial bubble etc. are recognized in the past
For the thing for being " dance of god ".Such as social physics researchers pass through various aspects in comprehensive collection human lives
Data crumbs analyze human behavior rule；Human behavior kineticist is attempted deep using quantification Spatiotemporal Statistical Analysis and modeling
Level understands public sentiment and transmission principle, optimization traffic programme, information recommendation etc..
Further in psychometrics, the psychological health states assessment to student is all the psychological row by individual students
It is measured indirectly for characterization.The over behavior performance of individual is dominated and is influenced by its psychological condition, and the difference of behavior can
To detect the difference of psychologic status, it is possible to scientific by mental health progress of the Psychology and behavior characterize data to student,
Accurately assessment.
It, not only can be with all of full-time comprehensive trace recording individual further under the support of modern information technologies fast development
External presentation, and either society or virtual society can use electronic data to realize whole seamless recording, this
Outside, big data storage management and cloud computing also provide the service of data efficient storage and analysis for Mental health evaluation.
Based on the above principles, according to one embodiment of present invention, as shown in Figure 1, to propose a kind of undergraduate psychological strong
Kang Yidong monitoring and warning and interfering system, the system include Virtual Space-entity space Psychology and behavior characterize data fusion
Layer, cloud storage platform, Psychology and behavior-behavior characterize data mapping model, Psychology and behavior unusual fluctuation monitoring and warning model, Psychology and behavior
Unusual fluctuation monitoring and warning software, the Virtual Space-entity space Psychology and behavior characterize data fused layer, cloud storage platform, the heart
Reason behavior-behavior characterize data mapping model, Psychology and behavior unusual fluctuation monitoring and warning model, Psychology and behavior unusual fluctuation monitoring and warning software
It is sequentially connected, the Psychology and behavior unusual fluctuation monitoring and warning software is connect with cloud storage platform.
As shown in figure 4, the Psychology and behavior unusual fluctuation monitoring and warning is specially with interfering system technology path
Step1: Virtual Space-entity space Psychology and behavior characterize data fused layer is established.
Integrated multiple information sources data, text information, image including acquisitions such as mobile Internet, information platform, Internet of Things
Information, relation data etc..
The text information is mainly derived from microblogging, information analysis its recent mood shape for being included by student's microblogging
Condition and mental health.It is main to pass through the removal microblogging texts such as complex form of Chinese characters conversion, interactive information filtering, text participle and part-of-speech tagging
The a large amount of noises for including in data utilize document frequency, information gain, subject analysis, emotion point on the basis of text representation
Analysis, keyword filtering, emoticon analysis etc. carry out text extraction, eventually form the microblogging text data of student.
Described image information is mainly derived from camera in classroom and acquires data and micro-blog photos etc..It is examined by using face
The technologies such as survey, Expression analysis, color analysis, scene analysis, time analysis extract the Sentiment orientation feature of image, then by its feature
It organically blends with other feature, ultimately forms the image data of student.
The relation data is mainly derived from school information platform, obtains the essential information of student on information platform, study
With job information, life information and virtual community information etc., to its data scrubbing, the purpose is to a variety of of self-information platform in future
Data carry out missing values processing, smooth noise data, identification or delete outlier, solve the problem of inconsistency of data, lack
Mistake value processing can by ignore tuple, be filled in manually missing values, using attribute center measurement fill in the modes such as missing values, noise
Can by branch mailbox, return, the methods of the point analysis that peels off remove.Ultimately form the relation data of student.
Above-mentioned text data, image data, relation data are further done into data preprocessing operation, including data set
At helping to reduce the redundancy and inconsistent in data set, facilitate the purpose is to merge the data from the storage of multiple data
The accuracy and speed for improving mining process thereafter can eliminate redundant data by correlation analysis during data integration；Number
According to specification, the purpose is to be used to obtain the expression of the specification of data set, the size of data is reduced, but still close to holding initial data
Integrality, data regularization can be carried out by the methods of dimension reduction, quantity reduction, data compression, wavelet transformation, PCA.According to this number
Data preprocess integrated framework realizes the mental representation data fusion of Virtual Space and entity space.
Step2: pretreated behavioural information data and its characteristic attribute are uploaded to cloud storage service platform.The cloud is deposited
The data for storing up service platform storage include following data: student's essential information, student's study and job information, student life are believed
Breath, student's virtual community information, figure acquisition information, student's sds self rating depression survey table information and factor score, student beck suppression
It is strongly fragrant to evaluate and test table and information certainly from evaluation and test table information and factor score, student sas anxiety.The virtual community information includes microblogging text
Word information, photographic intelligence, metadata and school share social platform text information, photographic intelligence, metadata etc..The cloud is deposited
Storage service platform is the self-built cloud storage service device of school, with high security, the advantages such as storage efficiency is high, access facilitates.
Step3: as shown in Fig. 2, establishing Psychology and behavior-behavior characterize data mapping model.Extract students psychology behavior number
The behavior characterize data relevant to Psychology and behavior in further eliminates the correlation between feature, reduces useless in feature
Information improves the value density of sample.The students psychology behavioral data include student's essential information, student and job information,
Life information and virtual community information etc..The students psychology behavior includes learning behavior, dietary behavior, motor behavior, online
Behavior etc..
Further, acquisition data dimension is reduced by PCA scheduling algorithm, utilizes covariance and the correlation of covariance matrix point
Determinant attribute relevant to Psychology and behavior in behavioral data is extracted in analysis, solves multiple dimensioned, more granularities, more of Psychology and behavior data
The problems such as source isomery.The determinant attribute extracted includes students' genders, the age, obtains scholarship situation, the situation that bears punishment, course
Excellent rate, pass rate, rate of failing, the outstanding door number of course and lattice door number, too late lattice door number, campus card consumption, disengaging place
Give up time, number of fetching water day, breakfast meal time, lunch meal time, dinner meal time, social information etc..
Step4: as shown in figure 3, establishing Psychology and behavior unusual fluctuation monitoring and warning model, polynary isomery big data ring is served
Border proposes the big data analysis frame of the fusion of multi-layer information mode based on data Layer, characteristic layer and model layer.
Further, with effective feature selecting algorithm, unrelated in all kinds of behavioral datas and redundancy feature is removed, is retained
Character subset related with mental health.The problem of being unevenly distributed weighing apparatus in view of assessment categories each in real data, from filtering type,
Packaging type, embedded multiple angles set out and carry out feature selecting, are that the multi-source multi-modal data in information system selects key
Feature obtains the specific symptom of assessment, prediction and early warning.
The feature selecting algorithm mainly includes Filter method, and certain weight, weight are assigned to every one-dimensional characteristic
The significance level of the dimensional feature is represented, is then ranked up according to weight size, main method is Chi-square Test, information increasing
Benefit, related coefficient etc.；Wrapper method sees subset selection as search optimization problem, generates different combination and other
Combination is compared, and finds out the best subset of effect；Embedded method, it is most significant to training pattern the purpose is to extract
Attribute, be added regular terms be most common method.
It is based further on Psychology and behavior-behavior characterize data mapping model, by deep learning algorithm, establishes and is suitable for being good for
Health information evaluation and predetermined deep learning model, level, each layer node number, excitation function including model etc..With depth mould
The superpower computing capability of type learns medical diagnosis of the mankind to single group data in conjunction with huge data source under big data environment
Ability, and play the excavation integration ability of machine learning scale of construction data big for multi-source.
The machine learning algorithm Selection Framework is as shown in figure 5, its main method is according to data set feature from a series of
The modeling method of best performance is selected in machine learning algorithm.On the basis of data prediction, selection is related to text, image etc.
The data set of aspect as training set and test set, will measure the multifarious number of tags of data set, label radix, label densities,
The indexs such as sample number are exported as mode input using the evaluation index of measure algorithm superiority-inferiority as model, including are based on sample
Evaluation index, the evaluation index based on label and the evaluation index based on sequence；By feature extraction, feature selecting to two classes
Index carries out dimensionality reduction, reduces model redundancy, then constructs regression model using SVR scheduling algorithm, and by above-mentioned data set to mould
Type is trained, to construct machine learning algorithm frame, improves algorithms selection efficiency.
Further accuracy rate and recall rate are currently used prediction model evaluation indexes, are that accurate evaluation Psychology and behavior is different
Dynamic monitoring and warning model accuracy.Present invention combination psychological knowledge proposes to measure model using false negative rate and false positive rate
Superiority and inferiority.False negative rate is the percentage for actually having Psychology and behavior unusual fluctuation or the positive, but being judged to no Psychology and behavior unusual fluctuation or feminine gender
Than reflection Screen test fails to pinpoint a disease in diagnosis the situation of Psychology and behavior unusual fluctuation individual；False positive rate is practical without Psychology and behavior unusual fluctuation or feminine gender,
But it is judged to the percentage of Psychology and behavior unusual fluctuation or the positive.The present invention excavates the Psychology and behavior unusual fluctuation monitoring and warning model established
It is expected that reach false positive lower than 50%, false negative levels off to 0.
The prediction model evaluation index specifically:
In formula, false negative number is actually to have Psychology and behavior unusual fluctuation or the individual of the positive to be identified as no Psychology and behavior unusual fluctuation
Or negative number, goldstandard positive number are the positive number that is diagnosed as, false negative rate is actually to have Psychology and behavior unusual fluctuation or sun
Property, but it is judged to the percentage of no Psychology and behavior unusual fluctuation or feminine gender；False positive number is actually to have Psychology and behavior without unusual fluctuation or yin
Property individual be identified as the number of Psychology and behavior unusual fluctuation or the positive, goldstandard feminine gender number is to be diagnosed as negative number, false
Positive rate is practical without Psychology and behavior unusual fluctuation or feminine gender, but is judged to the percentage of Psychology and behavior unusual fluctuation or the positive.
Step5: exploitation Psychology and behavior unusual fluctuation monitoring and warning software.
Based on Hadoop and Tensorflow big data analysis platform, above-mentioned theory research achievement is integrated, to educate big number
Based on, the Psychology and behavior unusual fluctuation monitoring and warning software towards college student group is researched and developed.It is existing " same that the software docks school
Heart cloud " platform, by cloud platform based on PaaS, micro services and technological development based on container realize data real-time acquisition and
Fusion, make it have with school other information system integration ability is strong, high reliablity, can the advantages such as elastic telescopic；Integrated Map-
The technologies such as Reduce, memory calculating, distributed data base, query engine, computing resource scheduling, flow data processing, realize data
Efficient process and analysis, provide Psychology and behavior unusual fluctuation Personal monitoring, early warning for students community, counsellor, psychological consultation center
With intervene etc. service.
Based on big data technology, the present embodiment Psychology and behavior unusual fluctuation monitoring and warning obtained has more quasi- with interfering system
Really, efficiently, comprehensively Mental health evaluation result and more accurate intervention stratege.
Claims (8)
1. a kind of undergraduate psychological behavior unusual fluctuation scheme Design method based on big data, which is characterized in that described is
System includes Virtual Space-entity space Psychology and behavior characterize data fused layer, cloud storage platform, Psychology and behavior-behavior characterization number
According to mapping model, Psychology and behavior unusual fluctuation monitoring and warning model, Psychology and behavior unusual fluctuation monitoring and warning software,
The Virtual Space-entity space Psychology and behavior characterize data fused layer, cloud storage platform, Psychology and behavior-behavior table
Sign data mapping model, Psychology and behavior unusual fluctuation monitoring and warning model, Psychology and behavior unusual fluctuation monitoring and warning software are sequentially connected, described
Psychology and behavior unusual fluctuation monitoring and warning software connect with cloud storage platform；
It is flat that the data that the Virtual Space-entity space Psychology and behavior characterize data fused layer obtains multiple information sources are stored in cloud
In platform, Psychology and behavior-behavior characterize data mapping model obtains the data being stored in cloud platform, passes through Psychology and behavior and row
Reasonable mapping model is established for the relationship of characterize data, it is pre- that master data and sample data are passed to Psychology and behavior unusual fluctuation monitoring
In alert model, so it is accurate, science, efficiently school students ' psychological health state is predicted；
The Psychology and behavior unusual fluctuation monitoring and warning software is embedded in said frame, mentions for students community, counsellor, psychological consultation center
For servicing accordingly.
2. a kind of undergraduate psychological behavior unusual fluctuation scheme Design method based on big data according to claim 1,
It is characterized in that, the Virtual Space-entity space Psychology and behavior characterize data fused layer content is including the use of microblogging crawler
The text information and image information of student are acquired,
The daily image information of student is acquired using image capturing system,
Essential information, study and work information, the life information etc. of student are acquired using school information platform,
The text information, image information, student information carry out final fusion, and are put into data prediction integrated framework
It is handled.
3. a kind of undergraduate psychological behavior unusual fluctuation scheme Design method based on big data according to claim 2,
It is characterized in that, the data prediction integrated framework specifically:
Step1: to from multiple information sources data carry out data scrubbing, by fill in missing values, smooth noise data, identification or
Delete the inconsistency that outlier solves data；
Step2: carrying out data integration, will get up and takes for all data acquisition systems of same individual from multiple data sources
Measure avoids redundancy when data integration, if it is necessary, Step1 can also be carried out again；
Step3: carrying out hough transformation, carries out simplifying expression to data set.
4. a kind of undergraduate psychological behavior unusual fluctuation scheme Design method based on big data according to claim 1,
It is characterized in that, the Psychology and behavior-behavior characterize data mapping model, including acquisition data are reduced by PCA scheduling algorithm
Dimension extracts crucial category relevant to Psychology and behavior in behavioral data using covariance and the correlation analysis of covariance matrix
Property, the determinant attribute extracted include students' genders, the age, obtain scholarship situation, the situation that bears punishment, course excellent rate and
The outstanding door number of lattice rate, rate of failing, course and lattice door number, too late lattice door number, campus card consumption, disengaging dormitory time, day
It fetches water number, breakfast meal time, lunch meal time, dinner meal time, social information etc..
5. a kind of undergraduate psychological behavior unusual fluctuation scheme Design method based on big data according to claim 1,
It is characterized in that, the Psychology and behavior unusual fluctuation monitoring and warning model includes big data analysis frame；
The big data analysis frame includes data Layer, characteristic layer and simulation layer, the data Layer, characteristic layer and simulation layer according to
Secondary connection.
6. a kind of undergraduate psychological behavior unusual fluctuation scheme Design method based on big data according to claim 5,
It is characterized in that, the big data analysis frame code fo practice specifically:
Step1: according to effective feature selecting algorithm, unrelated in all kinds of behavioral datas and redundancy feature is removed, retains and needs
Character subset；
Step2: Psychology and behavior-behavior characterize data mapping model, projected deep learning network learner are based on；Foundation is suitable for
The deep learning model of Mental health evaluation and prediction, level, each layer node number, activation primitive including model；
Step3: modeling all kinds of Psychology and behavior data characteristicses based on machine learning algorithm, carries out decision by various strategies
Grade fusion obtains total model, and improves model accuracy by Cooperative Optimization Algorithm, uses Increment Learning Algorithm for newly-increased data
Timely regularized learning algorithm model is to enhance the applicability of model.
7. a kind of undergraduate psychological behavior unusual fluctuation scheme Design method based on big data according to claim 1,
It is characterized in that, the Psychology and behavior unusual fluctuation monitoring and warning software is based on Hadoop and Tensorflow big data analysis platform
It is researched and developed, above-mentioned model and algorithm are realized in software insertion, specifically:
Step1: campus informatization system data (personnel system, card system, education administration system, Hospital are merged in data Layer
System, postgraduate's system, library etc.), sensing data (classroom camera data etc.), internet data (internet records, society
Hand over network behavior etc.) etc.；
Step2: decision-making level is based on above-mentioned model, and dynamically track finds Psychology and behavior unusual fluctuation individual；
Step3: in service layer to Psychology and behavior unusual fluctuation individual orient recommend psychological consultation article, daily work and rest planning, diet and
The information such as type of sports push Psychology and behavior unusual fluctuation individual information to students community, counsellor, psychological consultation center etc..
8. a kind of undergraduate psychological behavior unusual fluctuation scheme Design method based on big data according to claim 1,
It is characterized in that, technology path specifically:
Step1: Virtual Space-entity space Psychology and behavior characterize data fused layer is established
Integrated multiple information sources data, text information, image information including acquisitions such as mobile Internet, information platform, Internet of Things,
Relation data etc.；
The text information is mainly derived from microblogging, its recent emotional status of information analysis for being included by student's microblogging and
Mental health；It is main to pass through the removal microblogging text datas such as complex form of Chinese characters conversion, interactive information filtering, text participle and part-of-speech tagging
In include a large amount of noises, on the basis of text representation utilize document frequency, information gain, subject analysis, sentiment analysis, pass
Keyword filtering, emoticon analysis etc. carry out text extraction, eventually form the microblogging text data of student；
Described image information is mainly derived from camera acquisition data and micro-blog photos in classroom；By using Face datection, table
Mutual affection analysis, color analysis, scene analysis, time analysis technology extract image Sentiment orientation feature, then by its feature with it is other
Feature organically blends, and ultimately forms the image data of student；
The relation data is mainly derived from school information platform, obtains essential information, study and the work of student on information platform
Make information, life information and virtual community information, to its data scrubbing, the purpose is to a variety of data of self-information platform in future into
Outlier is deleted in the processing of row missing values, smooth noise data, identification, the problem of inconsistency of data is solved, at missing values
Reason by ignoring tuple, missing values are filled in manually, filling in missing values mode using the center measurement of attribute, noise by branch mailbox,
It returns, the removal of outlier analysis method；Ultimately form the relation data of student；
Above-mentioned text data, image data, relation data are further done into data preprocessing operation, including data integration,
Purpose is to merge the data from the storage of multiple data, helps to reduce the redundancy and inconsistent in data set, help to improve
Thereafter the accuracy and speed of mining process can eliminate redundant data by correlation analysis during data integration；Data rule
About, the purpose is to be used to obtain the expression of the specification of data set, the size of data is reduced, but still close to the complete of holding initial data
Property, data regularization can be carried out by dimension reduction, quantity reduction, data compression, wavelet transformation, PCA method；Located in advance according to this data
Integrated framework is managed, realizes the mental representation data fusion of Virtual Space and entity space；
Step2: pretreated behavioural information data and its characteristic attribute are uploaded to cloud storage service platform
The data of the cloud storage service platform storage include following data: student's essential information, student's study and job information,
Student life information, student's virtual community information, figure acquisition information, student's sds self rating depression survey table information and factor score,
Student's beck self rating depression surveys table information and factor score, student sas anxiety evaluation and test table and information certainly；The virtual community information
Social platform text information, photographic intelligence, metadata are shared including microblogging text information, photographic intelligence, metadata and school；
The cloud storage service platform is the self-built cloud storage service device of school；
Step3: Psychology and behavior-behavior characterize data mapping model is established
Behavior characterize data relevant to Psychology and behavior in students psychology behavioral data is extracted, the phase between feature is further eliminated
Guan Xing reduces the garbage in feature, improves the value density of sample；The students psychology behavioral data includes that student is basic
Information, student and job information, life information and virtual community information；The students psychology behavior includes learning behavior, diet
Behavior, motor behavior, internet behavior；
Further, acquisition data dimension is reduced by PCA algorithm to mention using covariance and the correlation analysis of covariance matrix
It takes determinant attribute relevant to Psychology and behavior in behavioral data, solves multiple dimensioned, more granularities, multi-source heterogeneous of Psychology and behavior data
Problem；The determinant attribute extracted include students' genders, the age, obtain scholarship situation, the situation that bears punishment, course excellent rate,
The outstanding door number of pass rate, rate of failing, course and lattice door number, not as good as lattice door number, campus card consumption, the disengaging dormitory time,
Day fetches water number, breakfast meal time, lunch meal time, dinner meal time, social information；
Step4: establishing Psychology and behavior unusual fluctuation monitoring and warning model, serves polynary isomery big data environment, proposes to be based on data
The big data analysis frame of the fusion of multi-layer information mode of layer, characteristic layer and model layer；
Further, with effective feature selecting algorithm, unrelated in all kinds of behavioral datas and redundancy feature, reservation and the heart are removed
The related character subset of reason health；The problem of being unevenly distributed weighing apparatus in view of assessment categories each in real data, from filtering type, packaging
Formula, embedded multiple angles set out and carry out feature selecting, are that the multi-source multi-modal data in information system selects key feature,
Obtain the specific symptom of assessment, prediction and early warning；
Further accuracy rate and recall rate are currently used prediction model evaluation indexes, are supervised for the unusual fluctuation of accurate evaluation Psychology and behavior
Survey Early-warning Model precision；In conjunction with psychological knowledge, the superiority and inferiority that model is measured using false negative rate and false positive rate is proposed；False yin
Property rate be actually have Psychology and behavior unusual fluctuation or the positive, but be judged to no Psychology and behavior unusual fluctuation or feminine gender percentage, reflect screening
The situation of Psychology and behavior unusual fluctuation individual is failed to pinpoint a disease in diagnosis in test；False positive rate is practical without Psychology and behavior unusual fluctuation or feminine gender, but has been judged to
The percentage of Psychology and behavior unusual fluctuation or the positive；
The prediction model evaluation index specifically:
In formula, false negative number is actually to have Psychology and behavior unusual fluctuation or the individual of the positive to be identified as no Psychology and behavior unusual fluctuation or yin
Property number, goldstandard positive number is to be diagnosed as positive number, false negative rate be actually have Psychology and behavior unusual fluctuation or the positive, but
It is judged to the percentage of no Psychology and behavior unusual fluctuation or feminine gender；False positive number is for actually having Psychology and behavior without unusual fluctuation or feminine gender
Body is identified as the number of Psychology and behavior unusual fluctuation or the positive, and goldstandard feminine gender number is to be diagnosed as negative number, false positive rate
It is reality without Psychology and behavior unusual fluctuation or feminine gender, but is judged to the percentage of Psychology and behavior unusual fluctuation or the positive
Step5: exploitation Psychology and behavior unusual fluctuation monitoring and warning software
Based on Hadoop and Tensorflow big data analysis platform, software docks school and has " concentric cloud " platform, passes through base
Realize that the real-time of data is acquired and merged in the cloud platform of PaaS, micro services and technological development based on container；Integrated Map-
Reduce, memory calculating, distributed data base, query engine, computing resource scheduling, flow data processing technique, realize data
Efficient process and analysis, for students community, counsellor, psychological consultation center provide Psychology and behavior unusual fluctuation Personal monitoring, early warning and
Intervene service.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810824392.1A CN109192310A (en) | 2018-07-25 | 2018-07-25 | A kind of undergraduate psychological behavior unusual fluctuation scheme Design method based on big data |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810824392.1A CN109192310A (en) | 2018-07-25 | 2018-07-25 | A kind of undergraduate psychological behavior unusual fluctuation scheme Design method based on big data |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109192310A true CN109192310A (en) | 2019-01-11 |
Family
ID=64936924
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810824392.1A Pending CN109192310A (en) | 2018-07-25 | 2018-07-25 | A kind of undergraduate psychological behavior unusual fluctuation scheme Design method based on big data |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109192310A (en) |
Cited By (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109785205A (en) * | 2019-01-24 | 2019-05-21 | 重庆电子工程职业学院 | It learns feelings and determines method, apparatus and intelligent terminal |
| CN110210559A (en) * | 2019-05-31 | 2019-09-06 | 北京小米移动软件有限公司 | Object screening technique and device, storage medium |
| CN110957039A (en) * | 2019-12-16 | 2020-04-03 | 河南科技学院 | Campus psychological coaching method and device based on deep learning |
| CN110993102A (en) * | 2019-11-18 | 2020-04-10 | 温州医科大学 | Campus big data-based student behavior and psychological detection result accurate analysis method and system |
| CN111477329A (en) * | 2020-03-31 | 2020-07-31 | 北京智能工场科技有限公司 | Method for evaluating psychological state based on image-text combination |
| CN111599472A (en) * | 2020-05-14 | 2020-08-28 | 重庆大学 | Method and device for recognizing psychological states of students and computer |
| CN111986781A (en) * | 2020-08-24 | 2020-11-24 | 龙马智芯（珠海横琴）科技有限公司 | Psychological treatment method and device based on man-machine interaction and user terminal |
| CN112380265A (en) * | 2020-10-16 | 2021-02-19 | 华南理工大学 | Psychological assessment data processing method, device and system and server |
| CN113093625A (en) * | 2021-04-12 | 2021-07-09 | 广州宏途教育网络科技有限公司 | Student behavior analysis system for intelligent classroom |
| CN113205869A (en) * | 2021-04-29 | 2021-08-03 | 华东师范大学 | Mental health detection and intervention method based on user social software head portrait |
| CN113436737A (en) * | 2021-06-24 | 2021-09-24 | 杭州师范大学 | Prediction evaluation method and device for depression level of large population |
| CN113539431A (en) * | 2021-07-16 | 2021-10-22 | 杨思思 | Social mental health education and career planning information big data platform |
| CN113609779A (en) * | 2021-08-16 | 2021-11-05 | 深圳力维智联技术有限公司 | Modeling method, device and equipment for distributed machine learning |
| CN114974586A (en) * | 2022-05-31 | 2022-08-30 | 六合熙诚(北京)信息科技有限公司 | Automatic monitoring method and system for artificial intelligence scale |
| CN115035974A (en) * | 2022-08-11 | 2022-09-09 | 北京科技大学 | Psychological assessment data management system and method |
| CN115274061A (en) * | 2022-09-26 | 2022-11-01 | 广州美术学院 | Interaction method, device, equipment and storage medium for soothing psychology of patient |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103905486A (en) * | 2012-12-26 | 2014-07-02 | 中国科学院心理研究所 | Mental health state evaluation method |
| CN107799165A (en) * | 2017-09-18 | 2018-03-13 | 华南理工大学 | A kind of psychological assessment method based on virtual reality technology |
| CN107895026A (en) * | 2017-11-17 | 2018-04-10 | 联奕科技有限公司 | A kind of implementation method of campus user portrait |
| CN107967572A (en) * | 2017-12-15 | 2018-04-27 | 华中师范大学 | A kind of intelligent server based on education big data |
| CN108121785A (en) * | 2017-12-15 | 2018-06-05 | 华中师范大学 | A kind of analysis method based on education big data |
-
2018
- 2018-07-25 CN CN201810824392.1A patent/CN109192310A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103905486A (en) * | 2012-12-26 | 2014-07-02 | 中国科学院心理研究所 | Mental health state evaluation method |
| CN107799165A (en) * | 2017-09-18 | 2018-03-13 | 华南理工大学 | A kind of psychological assessment method based on virtual reality technology |
| CN107895026A (en) * | 2017-11-17 | 2018-04-10 | 联奕科技有限公司 | A kind of implementation method of campus user portrait |
| CN107967572A (en) * | 2017-12-15 | 2018-04-27 | 华中师范大学 | A kind of intelligent server based on education big data |
| CN108121785A (en) * | 2017-12-15 | 2018-06-05 | 华中师范大学 | A kind of analysis method based on education big data |
Cited By (22)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109785205A (en) * | 2019-01-24 | 2019-05-21 | 重庆电子工程职业学院 | It learns feelings and determines method, apparatus and intelligent terminal |
| CN110210559B (en) * | 2019-05-31 | 2021-10-08 | 北京小米移动软件有限公司 | Object screening method and device and storage medium |
| CN110210559A (en) * | 2019-05-31 | 2019-09-06 | 北京小米移动软件有限公司 | Object screening technique and device, storage medium |
| CN110993102A (en) * | 2019-11-18 | 2020-04-10 | 温州医科大学 | Campus big data-based student behavior and psychological detection result accurate analysis method and system |
| CN110957039A (en) * | 2019-12-16 | 2020-04-03 | 河南科技学院 | Campus psychological coaching method and device based on deep learning |
| CN111477329A (en) * | 2020-03-31 | 2020-07-31 | 北京智能工场科技有限公司 | Method for evaluating psychological state based on image-text combination |
| CN111477329B (en) * | 2020-03-31 | 2021-04-13 | 北京智能工场科技有限公司 | Method for evaluating psychological state based on image-text combination |
| CN111599472A (en) * | 2020-05-14 | 2020-08-28 | 重庆大学 | Method and device for recognizing psychological states of students and computer |
| CN111599472B (en) * | 2020-05-14 | 2023-10-24 | 重庆大学 | Method and device for identifying psychological state of student and computer |
| CN111986781A (en) * | 2020-08-24 | 2020-11-24 | 龙马智芯（珠海横琴）科技有限公司 | Psychological treatment method and device based on man-machine interaction and user terminal |
| CN112380265A (en) * | 2020-10-16 | 2021-02-19 | 华南理工大学 | Psychological assessment data processing method, device and system and server |
| CN113093625A (en) * | 2021-04-12 | 2021-07-09 | 广州宏途教育网络科技有限公司 | Student behavior analysis system for intelligent classroom |
| CN113205869A (en) * | 2021-04-29 | 2021-08-03 | 华东师范大学 | Mental health detection and intervention method based on user social software head portrait |
| CN113205869B (en) * | 2021-04-29 | 2022-10-04 | 华东师范大学 | Mental health detection and intervention method based on user social software head portrait |
| CN113436737A (en) * | 2021-06-24 | 2021-09-24 | 杭州师范大学 | Prediction evaluation method and device for depression level of large population |
| CN113539431A (en) * | 2021-07-16 | 2021-10-22 | 杨思思 | Social mental health education and career planning information big data platform |
| CN113609779A (en) * | 2021-08-16 | 2021-11-05 | 深圳力维智联技术有限公司 | Modeling method, device and equipment for distributed machine learning |
| CN113609779B (en) * | 2021-08-16 | 2024-04-09 | 深圳力维智联技术有限公司 | Modeling method, device and equipment for distributed machine learning |
| CN114974586A (en) * | 2022-05-31 | 2022-08-30 | 六合熙诚(北京)信息科技有限公司 | Automatic monitoring method and system for artificial intelligence scale |
| CN115035974A (en) * | 2022-08-11 | 2022-09-09 | 北京科技大学 | Psychological assessment data management system and method |
| CN115035974B (en) * | 2022-08-11 | 2022-11-15 | 北京科技大学 | Psychological assessment data management system and method |
| CN115274061A (en) * | 2022-09-26 | 2022-11-01 | 广州美术学院 | Interaction method, device, equipment and storage medium for soothing psychology of patient |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109192310A (en) | A kind of undergraduate psychological behavior unusual fluctuation scheme Design method based on big data | |
| Han et al. | Data mining: concepts and techniques | |
| Wang et al. | A novelty detection patent mining approach for analyzing technological opportunities | |
| Altaf et al. | Applications of association rule mining in health informatics: a survey | |
| CN109256192A (en) | A kind of undergraduate psychological behavior unusual fluctuation monitoring and pre-alarming method neural network based | |
| Lei et al. | A novel data-driven robust framework based on machine learning and knowledge graph for disease classification | |
| Chatterjee et al. | An automatic ontology-based approach to support logical representation of observable and measurable data for healthy lifestyle management: Proof-of-concept study | |
| Shen et al. | Discovering the potential opportunities of scientific advancement and technological innovation: A case study of smart health monitoring technology | |
| Zhang et al. | Knowledge graph-based image classification refinement | |
| CN110275907A (en) | A kind of book recommendation method, device, equipment and readable storage medium storing program for executing | |
| Hachuel et al. | Augmenting gastrointestinal health: a deep learning approach to human stool recognition and characterization in macroscopic images | |
| Yang et al. | MADIS: A multimedia-aided disaster information integration system for emergency management | |
| Chen et al. | Developing a more responsive radiology resident dashboard | |
| Wu et al. | Short-term memory neural network-based cognitive computing in sports training complexity pattern recognition | |
| CN106055875A (en) | Dermatoglyph analysis and processing apparatus based on big data | |
| Alvarez et al. | Application of the spreading activation technique for recommending concepts of well-known ontologies in medical systems | |
| Guo et al. | From spoken narratives to domain knowledge: Mining linguistic data for medical image understanding | |
| Wang et al. | Kernel similarity-based multigranulation three-way decision approach to hypertension risk assessment with multi-source and multi-level structure data | |
| Baigorri et al. | A medical knowledge management system based on expert tagging (MKMST) | |
| Burstein et al. | Multivariate data-driven decision guidance for clinical scientists | |
| Zhao et al. | Logistic regression analysis of targeted poverty alleviation with big data in mobile network | |
| Wang et al. | Computer-Assisted Collaborative Learning for Enhancing Students Intellectual Ability Using Machine Learning Techniques | |
| Wei et al. | Medical College Education Data Analysis Method Based on Improved Deep Learning Algorithm | |
| Saraswat | Knowledge Discovery With Hybrid Data Mining Approach | |
| CN110853745A (en) | Skin disease patient standardization system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190111 |