CN110209709A - A method of concern human behavior analysis - Google Patents
A method of concern human behavior analysis Download PDFInfo
- Publication number
- CN110209709A CN110209709A CN201910490060.9A CN201910490060A CN110209709A CN 110209709 A CN110209709 A CN 110209709A CN 201910490060 A CN201910490060 A CN 201910490060A CN 110209709 A CN110209709 A CN 110209709A
- Authority
- CN
- China
- Prior art keywords
- personnel
- place
- data
- label
- concern
- 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
- 238000000034 method Methods 0.000 title claims abstract description 39
- 238000004458 analytical method Methods 0.000 title claims abstract description 31
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 16
- 239000000284 extract Substances 0.000 claims abstract description 5
- 238000004140 cleaning Methods 0.000 claims abstract description 4
- 230000006399 behavior Effects 0.000 claims description 23
- 238000010801 machine learning Methods 0.000 claims description 9
- 238000012549 training Methods 0.000 claims description 9
- 238000004891 communication Methods 0.000 claims description 8
- 230000000694 effects Effects 0.000 claims description 6
- 230000010354 integration Effects 0.000 claims description 4
- 238000002372 labelling Methods 0.000 claims description 4
- 230000002159 abnormal effect Effects 0.000 claims description 2
- 238000001914 filtration Methods 0.000 claims description 2
- 238000013075 data extraction Methods 0.000 claims 1
- PCHJSUWPFVWCPO-UHFFFAOYSA-N gold Chemical compound [Au] PCHJSUWPFVWCPO-UHFFFAOYSA-N 0.000 claims 1
- 239000010931 gold Substances 0.000 claims 1
- 229910052737 gold Inorganic materials 0.000 claims 1
- 230000003542 behavioural effect Effects 0.000 abstract description 3
- 238000007405 data analysis Methods 0.000 abstract description 3
- 239000003814 drug Substances 0.000 description 35
- 229940079593 drug Drugs 0.000 description 35
- 238000012545 processing Methods 0.000 description 5
- 238000000605 extraction Methods 0.000 description 4
- 239000002574 poison Substances 0.000 description 4
- 231100000614 poison Toxicity 0.000 description 4
- 238000011161 development Methods 0.000 description 3
- 230000002265 prevention Effects 0.000 description 3
- 241001269238 Data Species 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000007726 management method Methods 0.000 description 2
- 206010000117 Abnormal behaviour Diseases 0.000 description 1
- 208000001613 Gambling Diseases 0.000 description 1
- 238000010521 absorption reaction Methods 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 239000008280 blood Substances 0.000 description 1
- 210000004369 blood Anatomy 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000005352 clarification Methods 0.000 description 1
- 238000007635 classification algorithm Methods 0.000 description 1
- 230000001186 cumulative effect Effects 0.000 description 1
- 238000013506 data mapping Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 230000007717 exclusion Effects 0.000 description 1
- 230000002650 habitual effect Effects 0.000 description 1
- 230000008676 import Effects 0.000 description 1
- 238000000338 in vitro Methods 0.000 description 1
- 238000011835 investigation Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 238000005201 scrubbing Methods 0.000 description 1
- 238000012706 support-vector machine Methods 0.000 description 1
- 210000002700 urine Anatomy 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/21—Design, administration or maintenance of databases
- G06F16/215—Improving data quality; Data cleansing, e.g. de-duplication, removing invalid entries or correcting typographical errors
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/24—Querying
- G06F16/245—Query processing
- G06F16/2458—Special types of queries, e.g. statistical queries, fuzzy queries or distributed queries
- G06F16/2462—Approximate or statistical queries
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/10—Services
- G06Q50/26—Government or public services
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Strategic Management (AREA)
- Human Resources & Organizations (AREA)
- Data Mining & Analysis (AREA)
- General Engineering & Computer Science (AREA)
- Databases & Information Systems (AREA)
- Economics (AREA)
- Tourism & Hospitality (AREA)
- Software Systems (AREA)
- Marketing (AREA)
- Probability & Statistics with Applications (AREA)
- Quality & Reliability (AREA)
- Development Economics (AREA)
- Mathematical Physics (AREA)
- General Business, Economics & Management (AREA)
- Fuzzy Systems (AREA)
- Operations Research (AREA)
- General Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- Educational Administration (AREA)
- Game Theory and Decision Science (AREA)
- Entrepreneurship & Innovation (AREA)
- Health & Medical Sciences (AREA)
- Computational Linguistics (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Computation (AREA)
- Medical Informatics (AREA)
- Computing Systems (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The present invention relates to big data analysis technical fields, disclose a kind of method of concern human behavior analysis.Include: statistics locale data and extracts characteristic value;Statistician's data are simultaneously filtered cleaning, obtain the characteristic value of personnel;Sorting algorithm is used based on characteristic value, obtains the set and place set of doubtful criminal;It pays close attention to personnel characteristics' data and feature of place data carries out crash analysis, multidimensional data resonates to obtain the personnel that merit may occur and place;Similar label data generates the superposition of resonance characteristics integral or other treating methods realize the real-time dynamic integral of target, integrates given threshold according to this, carries out different grades of threshold value division, generate the warning information of different warning levels;And early warning situation is pushed to relevant departments and formulates treatment measures.Above scheme solves the management problems that personnel are paid close attention in current security administration, in conjunction with multidimensional data, in time, efficiently and accurately to the behavioural analysis of concern personnel.
Description
Technical field
The present invention relates to big data analysis technical field, especially a kind of method of concern human behavior analysis.
Background technique
There are many cases to occur daily in society, and the cases such as robbery, theft and gambling frequently occur.Currently, for these
Case, there are no outstanding prevention and handle before happening means, be substantially case occur and then come to case carry out processing and
It remedies.With the development of the social economy, movement of population quantity and range sharply increase, and urban population radix rapid growth, society
The frequency rapid development that case occurs, is frequently present of the case where criminal is habitual offender, and increasingly frequent, crime means occur for case
It becomes increasingly complex, traditional post and treatment measures are a kind of very big loss for society, and increase phase
The workload of staff is closed, it is more and more important in governance for the prevention and ex ante analysis of case.Internet, big number
According to and the technologies such as artificial intelligence fast development, provide the opportunity of change and innovation to pay close attention to the behavioural analysis of personnel.
Firstly, existing emphasis personnel managing and control system is mostly from data mapping or single means to the rail of emphasis personnel
Mark carries out analysis and studies and judges, and is difficult accurately to grasp the whereabouts of emphasis personnel, and find its abnormal behaviour in time.The big spininess of the prior art
Micro-judgment is carried out to track, acquisition of information source is single, to the behavior prediction inaccuracy of concern personnel, cannot timely find
With the generation of prevention merit.And often will cause erroneous judgement and misjudgement, security administration is influenced in the image of the public.
Summary of the invention
The technical problems to be solved by the present invention are: not enriching, being predicted not for information present in human behavior prediction
Accurately, the problems such as disposing not in time provides a kind of method of concern human behavior analysis.
The technical solution adopted by the invention is as follows: a kind of method of concern human behavior analysis, including following procedure:
Step S1 counts local all places, obtains the related basic data in place, passes through basic data
Characteristic information is extracted, it is labelled to known place sample, with the place of existing label come training machine learning algorithm, for not
The score value or probability of the place labelling and expression label confidence level known, there is one or more label in each place;
Step S2 obtains banking system, public security system, traffic system, the personnel for having charge sheet personnel in communication system
Data and the data of normal person;
Step S3, is filtered demographic data and clears up, and removes noise, obtains and case personnel or case place phase
The critical data of pass, extracting characteristic value training machine learning algorithm for each concern personnel further according to relevant critical data is
There is no the personnel of label to label and has the integral or probability for indicating label confidence level, everyone has one or more labels,
And there is corresponding label to integrate;
Step S4 uses sorting algorithm based on characteristic value, and to concern, personnel classify, and obtains tape label (with integral)
Personnel's set;
Step S5 uses sorting algorithm based on characteristic value, classifies to local place, obtains tape label (band product
Point) place set;
Step S6, set and place set to target person carry out the matching analysis, if the personal information of banking system
There is the personal information of Transaction Information or public security system criminal of having in locality of a crime set with the place in locality of a crime set
The demographic data that crime records perhaps traffic system records personnel to the place or communication system crossed in locality of a crime set
Personal information has the record for having call with the place in locality of a crime set, then pays close attention to personnel and pay close attention to the label data production in place
Raw resonance, similar label generate association, are integrated by algorithm to label and carry out resonance data, resonance integral superposition, and setting is abnormal
Threshold condition, if integral it is superimposed be integrated to given threshold triggering early warning, filter out the concern people that merit may occur
Member and concern place.Personnel and its current environment to concern carry out data collision, pass through people tag and site tabs
And environmental labels, current monitor can be understood more intuitively in the real-time trend labeling for the personnel that pay close attention to, business personnel by us
Target and the ambient conditions locating for it, concern clarification of objective label generate resonance data with people, the place etc. in scene.Data
Resonance is that the people of same label integrates generation association, obtains one of a people, thing, object and place according to the size of resonance integral
Real-time resonance integral, this integral can be used as the judgement basis whether target may have illegal tendency.
Further, the method for the concern human behavior analysis further includes the row to the concern personnel that merit may occur
The process being classified for probability: two or more labels of the same category form resonance data, and the integral of resonance is overlapped;
According to the main body for generating resonance, predicts possible criminal offence, in conjunction with current time or space attribute, be inferred to a suspect
Time and the place of criminal activity can be can be carried out;And the time of the criminal activity of prediction and ground are clicked through according to superimposed integral
The marking of row probability, is divided into different warning grades.
Further, different warning grade methods is divided are as follows: setting is different from low to high within the scope of range of integration
Integral threshold, arrive range of integration range for superimposed integral is regular, the integral when regular after is greater than different integral thresholds
Value, is judged as different warning grades.
Further, the method for the concern human behavior analysis further includes following procedure: being divided according to a suspect
Different warning grades, which is sent to related work system, formulates counter-measure.
Further, the machine learning model of label determined using classification or cluster, model passes through a certain number of
Positive sample and negative sample are trained to obtain.Sorting algorithm uses svm classifier algorithm or neural network classification algorithm.
Further, in the step S1, the method for characteristic information is extracted including but not limited to such as each place
Lower method: occurring the number of merit in specific time where Statistical Fields, the number of criminal occurs, and marking bound is arranged,
It is given a mark according to statistics number.
Further, in the step S3, the method for filtering and cleaning includes but is not limited to following method: for communication
The demographic data of system, including both call sides phone, both call sides ID number, name air time, call base station code, call
Time span;Demographic data for banking system include trade double hair names, transaction amount, exchange hour, transaction atm machine
Or bank outlets position and code transaction both sides' account;It include personnel concerning the case's name, case type for public security system data;
The track data of personnel is left behind for traffic system
Further, in the step S3, the method for extracting characteristic value includes but is not limited to following method: counting and has
The talk times of the personnel of previous conviction, there is the quantity of previous conviction personnel in ticket, statistician and case-involving place or relate to
The trading number of case personnel;The case-involving number of statistician, statistics removed the number in case-involving place, were obtained based on statistical value
The characteristic information of personnel carries out labeling to target, and obtain expression by model by characteristic information training label model
The probability value of a possibility that current label, using this value as integral basis.
Compared with prior art, by adopting the above technical scheme have the beneficial effect that the present invention combine big data technology, solve
The management problems that personnel are paid close attention in current security administration, in conjunction with multidimensional data, in time, efficiently to the behavioural analysis of concern personnel
With it is accurate.For under specific concern personnel and specified conditions behavior prediction and analysis, can effectively prevent case, efficiently
Control a suspect, technical guarantee is provided to security administration and governance.
Detailed description of the invention
Fig. 1 is that concern personnel of the present invention predict flow diagram.
Specific embodiment
The present invention is described further below.
Below by taking suspicious drug addict as an example, to the method for concern human behavior analysis of the invention.
Arresting for hidden poison personnel is extremely difficult, may just be metabolized out in vitro after a few houres to not if gas absorption dosage is small
Drugs index can be detected from urine or blood.So needing to be grabbed when drug addict organizes and takes drugs or taking drugs
It obtains.
The invention discloses a set of machine learning algorithm, may infer that whether a personnel are hidden malicious personnel, to hidden poison
Personnel are monitored, binding time and space attribute, infer hidden malicious human behavior, according to inferring preset processing scheme, such as it is hidden
Malicious personnel simultaneously or when successively appearing in same hotel, can take the processing means such as interim exclusion, to it with the personnel that are involved in drug traffic
It is interrogated and examined and is handled.As shown in Figure 1:
(1) basic data acquisition:
There are many basic data source, and all data that target can produce can be basic data, and the present embodiment uses
Be call bill data in communication system, the financial billing data in banking system, the hotel stay data in traffic system, mutually
Network acquires Internet bar's Internet data in networked system, and social unit, credit can also be acquired etc..
(2) after being collected into basic data, data are cleaned, filter invalid data and some unwanted fields, needle
To different data classifications, such as in ticket, some information related with operator, sim Card Type, call jingle bell are rejected
Number etc..Control field is added, for example ticket imports time, operator etc., suspicious extraction call caller and called name,
Air time.
The cleaning of other basic datas is also similar, rejects unwanted field, new control field is added.
(3) after data scrubbing finishes, we start to extract characteristic value.
The extracting method of characteristic value has very much, and most common is exactly to carry out signature analysis according to existing drug addict, has
For extraction characteristic value.The present invention carries out characteristics extraction to personnel and place simultaneously, and characteristic value is based on statistical value, parameter power
Weight can be back-calculated to obtain weight matrix by carrying out coefficient according to svm algorithm, carry out integral calculation based on characteristic value.
The characteristics extraction in place: case and there is the number for the personnel of committing a crime in place.Such as: characteristic value 1 is nearly one
It is involved in drug traffic in a month personnel's frequency of occurrence;The nearly month criminal's drug case part frequency of characteristic value 2.
The present embodiment collection is already recorded in the basic datas such as ticket, bill, lodging and the online of drug addict of case, makees
For sample.
Drug addict, the characteristic value point that we extract from ticket have:
1, it is involved in drug traffic in ticket personnel amount
2, high-risk areas is conversed
3, high-risk base station call
4, it converses after midnight
Characteristic value can have very much, unlimited above several.
Such as characteristic value: age, gender, occupation;Age characteristics value is exactly the age, and sex character value can use 0,1, occupation
Characteristic value can voluntarily regulation be replaced with digital code, and parameter area is unlimited, and numerical value, which finally can all be done, returns 1 change to handle.
Each personnel obtain a characteristic value for the data of different types of source, are formed based on different characteristic values multiple
Label with integrating properties;Because each personnel have multiple labels, each label has corresponded to different integrating properties.It is each
A label is exactly a portrait, and label is abundanter, and portrait is more plentiful, we are more intuitive to the awareness and understanding of personnel.
Each place obtains a characteristic value for the data of different types of source, is formed based on different characteristic values multiple
Label with integrating properties;Because each place has multiple labels, each label has corresponded to different integrating properties.It is each
A label is exactly a portrait, and label is abundanter, and portrait is more plentiful, we are more intuitive to the awareness and understanding in place.
(4) we use the SVM support vector machines of machine learning, and training one can be according to characteristic value to drug addict point
The classifier of class.
By existing sample, using characteristic value as feature vector, the model for drug addict is trained.
After having model, we can classify to doubtful drug addict, obtain the set of doubtful drug addict, obtain
To personnel's entity.
(5) personnel can classify, and place can also classify.
Firstly, count to local all places, the caseload that occurred according to place or there is crime
The number of personnel extracts characteristic value for each place, obtains place entity.
Such as hotel, hotel, we get the moving in after data of hotel, clean to data, obtain and data
Excavate the biggish data field of correlation.
According to existing data, we move in the background of personnel according to moving in and the check-out time, come analyze in hotel whether
The personnel that are involved in drug traffic often are received, the personnel that are involved in drug traffic enter and leave time and the space characteristics at hotel etc..Such as drug addict, always come in the morning
Open hourly paid hotel room etc..
According to existing hotel data, and the hotel for clearly thering is molecule of being involved in drug traffic often to enter and leave, we extract it is some enter
Firmly and quit the subscription of the characteristic values of data, such as the number that the personnel that are involved in drug traffic (and other types personnel concerning the case) move in, the time moved in
Range etc., the number of case of being involved in drug traffic (and other cases) obtain the characteristic value in the place, one hotel that is involved in drug traffic of training
Classifier classifies to the hotel that do not classify with this classifier, obtains the set at the doubtful hotel that is involved in drug traffic.
Place entity and personnel's entity are constituted data bins with the corresponding historical data of entity, integration data, case
Library.
(6) data are analyzed: it is based on data warehouse, the set and locality of a crime set to criminal carry out the matching analysis,
The demographic data in the set of criminal is extracted, carries out crash analysis, set concern target position with the data in the set of place
The data such as son, ticket, bill, for example, if the field All Activity in the personal information of banking system and locality of a crime set is believed
Cease demographic data of the personal information of perhaps public security system in locality of a crime set with previous conviction or traffic system
The personal information in record personnel to the place or communication system crossed in locality of a crime set has and the field in locality of a crime set
The record conversed, then successful match, carries out behavior prediction and filters out the concern personnel that merit may occur and concern place.Example
The people of a doubtful drug addict is paid close attention to as worked as us, in locality of a crime set sensitive hotel is suddenly appeared in and (relates to
It is malicious or doubtful be involved in drug traffic), which will trigger one and doubt criminaloid warning, if a doubtful drug addict, appearance
Sensitive hotel has had drug addict other in the set of criminal to move in, and the room opened is adjacent or does not rent a room in a hotel straight
It taps into hotel, then successful match, judgement may have the crime being involved in drug traffic.By the step (2) of the result update of data analysis
The process of middle data cleansing processing.
Based on above-mentioned matching result, investigation scheme is formulated, early warning is pushed to the related management personnel of surrounding, notice connects
It receives personnel targets situation and doubts criminaloid details.Because the prediction of the application is drawn a portrait and resonance data early warning based on feature,
The more accurate image of early warning situation of push provides treatment measures reference for business personnel, formulates early warning disposal method with this.Example
It such as organizes the selective examination at a hotel or interrogates and examines criminal and doubtful criminal perhaps to investigate or directly carry out
It interrogates and examines and poison checks.The data of screening are allowed to generate value.
Probability marking is carried out to matched concern personnel and concern place: being to pay close attention to the sum of the characteristic value of personnel and place
One integral label, it will be assumed that the lower limit of characteristic value is 0 point, and the upper limit is 100 points.When this person be involved in drug traffic characteristic value label be 80
Point, the label 90 of being involved in drug traffic of feature of place value divides, while the target paid close attention to is not necessarily individual personnel or scene, the embodiment
There are 4 personnel to be involved in drug traffic herein floor activity mistake, other two personnel that are involved in drug traffic are had also appeared in scene, integral is also 80-90 respectively,
Place class label of being involved in drug traffic integrates the label of being involved in drug traffic of 90. this four people and generates resonance, and corresponding integral is cumulative to obtain relating to for 340-360
Poison integral, we are to this Integral Processing, the range of specification to 0-100, and the integral after specification is 85-90, if in the setting of this range
Dry warning grade, such as 60-70 blue early warning, 70-80 yellow early warning 80-90 orange warning 90-100 red early warning are above-mentioned to set
It sets successively backward, warning grade is higher.
The invention is not limited to specific embodiments above-mentioned.The present invention, which expands to, any in the present specification to be disclosed
New feature or any new combination, and disclose any new method or process the step of or any new combination.If this
Field technical staff is altered or modified not departing from the unsubstantiality that spirit of the invention is done, should belong to power of the present invention
The claimed range of benefit.
Claims (8)
1. a kind of method of concern human behavior analysis, which is characterized in that including following procedure:
Step S1 counts local all places, obtains the related basic data in place, is extracted by basic data
Characteristic information, it is labelled to known place sample, it is unknown with the place of existing label come training machine learning algorithm
Place labelling and the score value or probability for indicating label confidence level, there is one or more label in each place;
Step S2 obtains banking system, public security system, traffic system, the demographic data for having charge sheet personnel in communication system
And the data of normal person;
Step S3, is filtered demographic data and clears up, and removes noise, obtains relevant to case personnel or case place
Critical data, extracting characteristic value training machine learning algorithm for each concern personnel further according to relevant critical data is not have
The personnel of label label and have the integral or probability that indicate label confidence level, everyone has one or more labels, and
There is corresponding label to integrate;
Step S4 uses sorting algorithm based on characteristic value, and to concern, personnel classify, and obtain the set of the personnel of tape label;
Step S5 uses sorting algorithm based on characteristic value, classifies to local place, obtain the place collection of tape label
It closes;
Step S6, set and place set to target person carry out the matching analysis, if the personal information of banking system and criminal
Place in guilty place set has the personal information of Transaction Information or public security system to have crime note in locality of a crime set
The demographic data for recording perhaps traffic system records personnel to the place crossed in locality of a crime set or the personnel of communication system
Information has the record for having call with the place in locality of a crime set, then pays close attention to personnel and pay close attention to the label data generation in place altogether
Vibration, similar label generate association, are integrated by algorithm to label and carry out resonance data, and resonance integral superposition sets abnormal threshold
Value condition, if integral it is superimposed be integrated to given threshold triggering early warning, filter out may occur merit concern personnel and
Pay close attention to place.
2. the method for concern human behavior analysis as described in claim 1, which is characterized in that the concern human behavior analysis
Method further include the process being classified to the behavior probability for the concern personnel that merit may occur: two of the same category or
Multiple labels form resonance data, and the integral of resonance is overlapped;According to the main body for generating resonance, possible crime row is predicted
For in conjunction with current time or space attribute, time and the place of criminal activity can be can be carried out by being inferred to a suspect;And according to
Superimposed integral carries out probability marking to the time of the criminal activity of prediction and place, is divided into different warning grades.
3. the method for concern human behavior analysis as claimed in claim 2, which is characterized in that divide different warning grade sides
Method are as follows: different integral thresholds is set from low to high within the scope of range of integration, superimposed integral is regular on integral
Lower range, the integral when regular after are greater than different integral thresholds, are judged as different warning grades.
4. the method for concern human behavior analysis as claimed in claim 3, which is characterized in that the concern human behavior analysis
Method further include following procedure: according to the different warning grades divided to a suspect, which is sent to phase
Work system is closed, counter-measure is formulated.
5. as described in claim 1 concern human behavior analysis method, which is characterized in that label determine using classification or
The machine learning model of person's cluster, model are trained to obtain by a certain number of positive samples and negative sample.
6. the method for concern human behavior analysis as claimed in claim 5, which is characterized in that in the step S1, pass through base
The method of plinth data extraction characteristic information: occurring the number of merit in specific time where Statistical Fields, time of criminal occurs
Number, the classification obtained by training or cluster machine learning algorithm obtain the probability of feature tag and feature tag, feature
Score value of the probability of label as feature tag.
7. as claimed in claim 6 concern human behavior analysis method, which is characterized in that in the step S3, filtering and
The method of cleaning are as follows: converse for the demographic data of communication system, including both call sides phone, both call sides ID number, name
Time, call base station code, air time length;Demographic data for banking system includes the double hair names, trade gold of transaction
Volume, exchange hour, the atm machine of transaction or bank outlets position and code transaction both sides' account;For public security system data packet
Include personnel concerning the case's name, case type;The track data of personnel is left behind for traffic system.
8. the method for concern human behavior analysis as claimed in claim 7, which is characterized in that in the step S3, extract special
The method of value indicative are as follows: count and have previous conviction personnel talk times, have the quantity of previous conviction personnel, statistics in ticket
The trading number of personnel and case-involving place or personnel concerning the case;The case-involving number of statistician, statistics went to case-involving place
Number, the characteristic information of personnel is obtained based on statistical value, by characteristic information training label model, label point is carried out to target
Class, and the probability value for a possibility that indicating current label is obtained by model, it is basic as integral using this value.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910490060.9A CN110209709A (en) | 2019-06-06 | 2019-06-06 | A method of concern human behavior analysis |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910490060.9A CN110209709A (en) | 2019-06-06 | 2019-06-06 | A method of concern human behavior analysis |
Publications (1)
Publication Number | Publication Date |
---|---|
CN110209709A true CN110209709A (en) | 2019-09-06 |
Family
ID=67791304
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910490060.9A Pending CN110209709A (en) | 2019-06-06 | 2019-06-06 | A method of concern human behavior analysis |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110209709A (en) |
Cited By (20)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110633915A (en) * | 2019-09-24 | 2019-12-31 | 北京明略软件系统有限公司 | High-risk place identification method and device |
CN110634093A (en) * | 2019-09-26 | 2019-12-31 | 四川科瑞软件有限责任公司 | Travel analysis method for virus-involved people |
CN110647561A (en) * | 2019-09-26 | 2020-01-03 | 四川科瑞软件有限责任公司 | Communication track analysis method for drug-involved personnel |
CN110659335A (en) * | 2019-09-26 | 2020-01-07 | 四川科瑞软件有限责任公司 | Analysis method for recessive drug addicts |
CN110704469A (en) * | 2019-10-22 | 2020-01-17 | 北京明略软件系统有限公司 | Updating method and updating device of early warning level and readable storage medium |
CN110852517A (en) * | 2019-11-15 | 2020-02-28 | 北京明略软件系统有限公司 | Abnormal behavior early warning method and device, data processing equipment and storage medium |
CN111080005A (en) * | 2019-12-12 | 2020-04-28 | 华中科技大学 | Support vector machine-based public security risk early warning method and system |
CN111310780A (en) * | 2019-12-18 | 2020-06-19 | 北京明略软件系统有限公司 | Police service processing method and device |
CN111310100A (en) * | 2020-01-06 | 2020-06-19 | 北京中天锋安全防护技术有限公司 | Intelligent evaluation method for drug addict |
CN111311056A (en) * | 2020-01-06 | 2020-06-19 | 北京中天锋安全防护技术有限公司 | Drug addict risk monitoring method |
CN111460246A (en) * | 2019-12-19 | 2020-07-28 | 南京柏跃软件有限公司 | Real-time activity abnormal person discovery method based on data mining and density detection |
CN112132357A (en) * | 2020-09-29 | 2020-12-25 | 佳都新太科技股份有限公司 | Behavior prediction method, behavior prediction device, behavior prediction equipment and storage medium based on big data |
CN112233480A (en) * | 2020-10-23 | 2021-01-15 | 重庆海知声科技有限公司 | Online interactive education system and method |
CN112396449A (en) * | 2020-06-30 | 2021-02-23 | 安徽听见科技有限公司 | Method, device, equipment and storage medium for predicting group activities |
CN112651433A (en) * | 2020-12-17 | 2021-04-13 | 广州锦行网络科技有限公司 | Abnormal behavior analysis method for privileged account |
CN112651442A (en) * | 2020-12-25 | 2021-04-13 | 南京中兴力维软件有限公司 | Crime prediction method, crime prediction device, crime prediction equipment and computer readable storage medium |
CN113313344A (en) * | 2021-04-13 | 2021-08-27 | 武汉烽火众智数字技术有限责任公司 | Label system construction method and system fusing multiple modes |
CN113656485A (en) * | 2021-01-30 | 2021-11-16 | 河南信安通信技术股份有限公司 | Figure data analysis method and system under big data |
CN114091844A (en) * | 2021-11-01 | 2022-02-25 | 山东心法科技有限公司 | Early warning method, device and storage medium for crime reoccurrence of violent personnel |
TWI807949B (en) * | 2022-08-02 | 2023-07-01 | 兆豐國際商業銀行股份有限公司 | Alert system and alert method of fraud prevention for financial consultant |
-
2019
- 2019-06-06 CN CN201910490060.9A patent/CN110209709A/en active Pending
Cited By (26)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110633915A (en) * | 2019-09-24 | 2019-12-31 | 北京明略软件系统有限公司 | High-risk place identification method and device |
CN110634093A (en) * | 2019-09-26 | 2019-12-31 | 四川科瑞软件有限责任公司 | Travel analysis method for virus-involved people |
CN110647561A (en) * | 2019-09-26 | 2020-01-03 | 四川科瑞软件有限责任公司 | Communication track analysis method for drug-involved personnel |
CN110659335A (en) * | 2019-09-26 | 2020-01-07 | 四川科瑞软件有限责任公司 | Analysis method for recessive drug addicts |
CN110704469A (en) * | 2019-10-22 | 2020-01-17 | 北京明略软件系统有限公司 | Updating method and updating device of early warning level and readable storage medium |
CN110704469B (en) * | 2019-10-22 | 2022-11-11 | 北京明智和术科技有限公司 | Updating method and updating device of early warning level and readable storage medium |
CN110852517A (en) * | 2019-11-15 | 2020-02-28 | 北京明略软件系统有限公司 | Abnormal behavior early warning method and device, data processing equipment and storage medium |
CN110852517B (en) * | 2019-11-15 | 2023-08-25 | 北京明智和术科技有限公司 | Abnormal behavior early warning method and device, data processing equipment and storage medium |
CN111080005A (en) * | 2019-12-12 | 2020-04-28 | 华中科技大学 | Support vector machine-based public security risk early warning method and system |
CN111080005B (en) * | 2019-12-12 | 2022-05-17 | 华中科技大学 | Support vector machine-based public security risk early warning method and system |
CN111310780A (en) * | 2019-12-18 | 2020-06-19 | 北京明略软件系统有限公司 | Police service processing method and device |
CN111460246A (en) * | 2019-12-19 | 2020-07-28 | 南京柏跃软件有限公司 | Real-time activity abnormal person discovery method based on data mining and density detection |
CN111460246B (en) * | 2019-12-19 | 2020-12-08 | 南京柏跃软件有限公司 | Real-time activity abnormal person discovery method based on data mining and density detection |
CN111310100A (en) * | 2020-01-06 | 2020-06-19 | 北京中天锋安全防护技术有限公司 | Intelligent evaluation method for drug addict |
CN111311056A (en) * | 2020-01-06 | 2020-06-19 | 北京中天锋安全防护技术有限公司 | Drug addict risk monitoring method |
CN112396449A (en) * | 2020-06-30 | 2021-02-23 | 安徽听见科技有限公司 | Method, device, equipment and storage medium for predicting group activities |
CN112132357A (en) * | 2020-09-29 | 2020-12-25 | 佳都新太科技股份有限公司 | Behavior prediction method, behavior prediction device, behavior prediction equipment and storage medium based on big data |
CN112233480A (en) * | 2020-10-23 | 2021-01-15 | 重庆海知声科技有限公司 | Online interactive education system and method |
CN112651433B (en) * | 2020-12-17 | 2021-12-14 | 广州锦行网络科技有限公司 | Abnormal behavior analysis method for privileged account |
CN112651433A (en) * | 2020-12-17 | 2021-04-13 | 广州锦行网络科技有限公司 | Abnormal behavior analysis method for privileged account |
CN112651442A (en) * | 2020-12-25 | 2021-04-13 | 南京中兴力维软件有限公司 | Crime prediction method, crime prediction device, crime prediction equipment and computer readable storage medium |
CN113656485A (en) * | 2021-01-30 | 2021-11-16 | 河南信安通信技术股份有限公司 | Figure data analysis method and system under big data |
CN113313344A (en) * | 2021-04-13 | 2021-08-27 | 武汉烽火众智数字技术有限责任公司 | Label system construction method and system fusing multiple modes |
CN113313344B (en) * | 2021-04-13 | 2023-03-31 | 武汉烽火众智数字技术有限责任公司 | Label system construction method and system fusing multiple modes |
CN114091844A (en) * | 2021-11-01 | 2022-02-25 | 山东心法科技有限公司 | Early warning method, device and storage medium for crime reoccurrence of violent personnel |
TWI807949B (en) * | 2022-08-02 | 2023-07-01 | 兆豐國際商業銀行股份有限公司 | Alert system and alert method of fraud prevention for financial consultant |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110209709A (en) | A method of concern human behavior analysis | |
Sathyadevan et al. | Crime analysis and prediction using data mining | |
CN105404890B (en) | A kind of criminal gang's method of discrimination for taking track space and time order into account | |
Rossy et al. | Integrating forensic information in a crime intelligence database | |
Ribaux et al. | Case based reasoning in criminal intelligence using forensic case data | |
US20090099884A1 (en) | Method and system for detecting fraud based on financial records | |
CN110298254A (en) | A kind of analysis method and system for personnel's abnormal behaviour | |
CN102222283A (en) | Data management system for face recognition | |
Adderley et al. | Police crime recording and investigation systems–A user’s view | |
CN110349372B (en) | Method and device for early warning abnormal activities of family aged care personnel | |
CN110390031A (en) | Information processing method and device, vision facilities and storage medium | |
CN107527313A (en) | User Activity mode division and attribute estimation method | |
CN110727805A (en) | Community knowledge graph construction method and system | |
CN105205048B (en) | A kind of hot word analytic statistics system and method | |
Zhao et al. | Detecting pickpocketing gangs on buses with smart card data | |
Manoj et al. | A Strategy for Identification and Prevention of Crime using various Classifiers | |
Grover et al. | Review of current crime prediction techniques | |
CN113628738B (en) | Active epidemic prevention device and method for intelligent recovery and detection of oronasal mask | |
Tarlekar et al. | Geographical Crime Rate Prediction System | |
Sundaram | Classification rules by decision tree for disease prediction | |
Barnett‐Ryan et al. | The dynamic nature of crime statistics | |
Adderley | The use of data mining techniques in crime trend analysis and offender profiling | |
CN111210208A (en) | Public security control of drug addicts | |
Andresen et al. | An evaluation of bylaw and policy changes on pharmacy robberies in British Columbia | |
Munasinghe et al. | Criminal short listing and crime forecasting based on modus operandi |
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 |
Application publication date: 20190906 |
|
RJ01 | Rejection of invention patent application after publication |