A kind of Crime Spreading complex network node analysis method, apparatus and user terminalTechnical field
The present invention relates to Complex Networks Analysis technical field, more specifically to a kind of Crime Spreading complex network knotPoint analysis method, apparatus and user terminal.
Background technology
Complex network (Complex Network), refers to self-organizing, self similarity, attractor, worldlet, uncalibrated visual servoIn partly or entirely property network.At present in Complex Networks Analysis, it is in the signature analysis of analysis complex network mostly,The aspect of the high level such as kinetic model.For in the method for crime analysis, there is also much from theoretical and experimental viewpoint phaseResearch is closed, but is still rested on based on questionnaire survey mostly, in the case of being subject to supplemented by the method for statistical analysis.
Although there are the analysis methods of Various Complex network now, still most of to be in excessively macroscopical, mainly forThe high-level feature of some of network is analyzed, such as some statistical natures of network：Whether small-world network, rule are metNetwork, network node number of types, average path length, cluster coefficients etc..And the analysis for single node is actually rare.
In short, the existing analysis for crime complex network, only rests on questionnaire survey or Basic Statistics and Tables methodOn, it can not accomplish the analysis for single node, so as to which the accurate analysis result for suspect can not be obtained, to caseFurther the analysis to the peripheral information of suspect and investigation bring great inconvenience to investigator.
Invention content
In view of this, the present invention provide a kind of Crime Spreading complex network node analysis method, apparatus and user terminal withSolve the deficiencies in the prior art.
To solve the above problems, the present invention provides a kind of Crime Spreading complex network node analysis method, including：
It is built and the suspicion according to the personal perimeter data of suspect and with the relation data of the suspect related personnelDoubt the corresponding dynamic criminal network of people；
Extract the corresponding destination node feature of the suspect in the dynamic criminal network and according to the target knotPoint feature establishes vector；
According to the destination node feature and the vector structure classification analysis model；
The state supervision to the destination node is carried out based on the classification analysis model.
Preferably, described " according to the destination node and the vector structure classification analysis model ", including：
In the dynamic criminal network, according to the vector of the destination node before and after crime, the mesh is obtainedMark the dynamic change characterization vector of node；
The crime vector of the suspect is built based on the dynamic change characterization vector；
According to the crime vector, classification analysis model is established.
Preferably, it is described " in the dynamic criminal network, according to the destination node before and after crime it is described toAmount obtains the dynamic change characterization vector of the destination node ", including：
In the dynamic criminal network, primary vector is built for destination node described before crime, for institute after crimeState destination node structure secondary vector；
The difference of the secondary vector and the primary vector is calculated, the dynamic change for obtaining the destination node is specialSign vector.
Preferably, the described " relation data according to the personal perimeter data of suspect and with the suspect related personnelStructure dynamic criminal network corresponding with the suspect ", including：
The network topological diagram of the different time node of the suspect is built according to the personal perimeter data；
In the network topological diagram, the suspect and suspect related personnel setting to different time nodeState relation label；
According to the network topological diagram for including the state relation label, the different time node of the suspect is builtThe dynamic criminal network.
Preferably, the destination node feature includes network statistics feature, network path feature, network Sub-Image Feature, spyDetermine service feature and the state relation label；
It is described " to extract the corresponding destination node feature of the suspect in the dynamic criminal network and according to the meshMark node feature and establish vector ", including：
According to default node degree, the network statistics in the dynamic criminal network in the default node degree is extractedFeature, the network path feature, the network Sub-Image Feature, the certain service features and the state relation label；
The network statistics feature, the network path feature, network in the dynamic criminal networkFigure feature, the certain service features and the state relation label establish the corresponding vector of the destination node.
Preferably, described " state supervision to the destination node is carried out based on the classification analysis model ", including：
The classification analysis model of the destination node and the vector are analyzed, obtains the different time of the suspectThe analysis result of node；
State supervision to the destination node is carried out according to the analysis result.
Preferably, the described " relation data according to the personal perimeter data of suspect and with the suspect related personnelBefore structure dynamic criminal network corresponding with the suspect ", further include：
Obtain the personal perimeter data of the suspect；Obtain the pass of relationship between the related personnel and the suspectCoefficient evidence.
In addition, to solve the above problems, the present invention also provides a kind of Crime Spreading network node analytical equipment, including：StructureModel block, extraction module and supervision module；
The structure module, for the personal perimeter data according to suspect and the relationship with the suspect related personnelData build dynamic criminal network corresponding with the suspect；
The extraction module, for extracting the corresponding destination node feature of the suspect in the dynamic criminal networkAnd vector is established according to the destination node feature；
The structure module is additionally operable to according to the destination node feature and the vector structure classification analysis model；
The supervision module, for carrying out the state supervision to the destination node based on the classification analysis model.
In addition, to solve the above problems, the present invention also provides a kind of user terminal, including memory and processor, instituteMemory is stated for storing Crime Spreading network node analysis program, the processor runs the Crime Spreading network node pointProgram is analysed so that the user terminal performs Crime Spreading complex network node analysis method as described above.
In addition, to solve the above problems, the present invention also provides a kind of computer readable storage medium, it is described computer-readableCrime Spreading network node analysis program is stored on storage medium, the Crime Spreading network node analysis program is by processorCrime Spreading complex network node analysis method as described above is realized during execution.
A kind of Crime Spreading complex network node analysis method, apparatus provided by the invention and user terminal.Wherein, this hairBright provided method is started with from destination node, the dynamic relationship network with reference to residing for the specific node, various dimensions structure networkFeature so as to analyze network topology change of the destination node before and after crime, is opened up so as to fulfill by the network of destination nodeThat flutters mutation analysis obtains the analysis result of suspect, and analysis result is accurate, be case investigator further to crimeThe analysis and investigation of the peripheral information of suspect bring great convenience.
Description of the drawings
Fig. 1 is the hardware running environment that Crime Spreading complex network node analysis embodiment of the method scheme of the present invention is related toStructure diagram；
Fig. 2 is the flow diagram of Crime Spreading complex network node analysis method first embodiment of the present invention；
Fig. 3 is the flow diagram of Crime Spreading complex network node analysis method second embodiment of the present invention；
Fig. 4 is the flow diagram of Crime Spreading complex network node analysis method 3rd embodiment of the present invention；
Fig. 5 is the flow diagram of Crime Spreading complex network node analysis method fourth embodiment of the present invention；
Fig. 6 is the flow diagram of the 5th embodiment of Crime Spreading complex network node analysis method of the present invention；
Fig. 7 is the flow diagram of Crime Spreading complex network node analysis method sixth embodiment of the present invention；
Fig. 8 is network of personal connections before the suspect of the 7th embodiment of Crime Spreading complex network node analysis method of the present invention takes drugsSchematic diagram；
Fig. 9 is network of personal connections after the suspect of the 7th embodiment of Crime Spreading complex network node analysis method of the present invention takes drugsSchematic diagram；
Figure 10 is the flow diagram of the 7th embodiment of Crime Spreading complex network node analysis method of the present invention；
Figure 11 is the high-level schematic functional block diagram of Crime Spreading network node analytical equipment of the present invention.
The embodiments will be further described with reference to the accompanying drawings for the realization, the function and the advantages of the object of the present invention.
Specific embodiment
The embodiment of the present invention is described below in detail, wherein same or similar label represents same or like from beginning to endElement or with same or like function element.
In addition, term " first ", " second " are only used for description purpose, and it is not intended that instruction or hint relative importanceOr the implicit quantity for indicating indicated technical characteristic.Define " first " as a result, the feature of " second " can be expressed orImplicitly include one or more this feature.In the description of the present invention, " multiple " are meant that two or more,Unless otherwise specifically defined.
In the present invention unless specifically defined or limited otherwise, term " installation ", " connected ", " connection ", " fixation " etc.Term should be interpreted broadly, for example, it may be being fixedly connected or being detachably connected or integral；Can be that machinery connectsIt connects or is electrically connected；It can be directly connected, can also be indirectly connected by intermediary, can be in two elementsThe connection in portion or the interaction relationship of two elements.It for the ordinary skill in the art, can be according to specific feelingsCondition understands the concrete meaning of above-mentioned term in the present invention.
It should be appreciated that the specific embodiments described herein are merely illustrative of the present invention, it is not intended to limit the present invention.
As shown in Figure 1, the structure diagram of the hardware running environment of terminal that Fig. 1, which is the embodiment of the present invention, to be related to.
Terminal of the embodiment of the present invention can be PC or smart mobile phone, tablet computer, E-book reader, MP3 are broadcastPutting device, MP4 players, pocket computer etc. has the packaged type terminal device of display function.
As shown in Figure 1, the terminal can include：Processor 1001, such as CPU, network interface 1004, user interface1003, memory 1005, communication bus 1002.Wherein, communication bus 1002 is used to implement the connection communication between these components.User interface 1003 can include display screen, input unit such as keyboard, remote controler, and optional user interface 1003 can also includeStandard wireline interface and wireless interface.Network interface 1004 can optionally include standard wireline interface and wireless interface (such asWI-FI interfaces).Memory 1005 can be high-speed RAM memory or the memory of stabilization, such as magnetic disk storage.Memory 1005 optionally can also be the storage device independently of aforementioned processor 1001.
Optionally, terminal can also include camera, RF (Radio Frequency, radio frequency) circuit, sensor, audioCircuit, WiFi module etc..In addition, mobile terminal can also configure gyroscope, barometer, hygrometer, thermometer, infrared ray sensingThe other sensors such as device, details are not described herein.
It will be understood by those skilled in the art that the restriction of the terminal shown in Fig. 1 not structure paired terminal, can include thanIt illustrates more or fewer components and either combines certain components or different components arrangement.
As shown in Figure 1, as operating system, number can be included in a kind of memory 1005 of computer readable storage mediumAccording to interface control program, network attachment procedure and Crime Spreading network node analysis program.
A kind of Crime Spreading complex network node analysis method, apparatus provided by the invention and user terminal.Wherein, it is describedMethod realization obtains the analysis result of suspect by what the network topology change of destination node was analyzed, and analysis result is accurateReally, for case investigator, further the analysis to the peripheral information of suspect and investigation bring great convenience.
Embodiment 1：
With reference to Fig. 2, first embodiment of the invention provides a kind of Crime Spreading complex network node analysis method, including：
Step S100 is built according to the personal perimeter data of suspect and with the relation data of the suspect related personnelDynamic criminal network corresponding with the suspect.
Above-mentioned, the suspicion artificially needs to be analyzed the target that is retrieved of investigation, suspect or other withThe related related personnel of this case.
It is above-mentioned, personal perimeter data, for can be collected on network with the relevant all data of suspect, can includeIt is related to the various aspects such as the clothing, food, lodging and transportion -- basic necessities of life of suspect, such as traffic data, personal information etc..
Above-mentioned, the suspect related personnel is and the relevant different crowd of suspect, such as relative, friend and suspicionPeople has the people of debt and credit relationship, the object clashed etc..
Above-mentioned, relation data can include relationship situation between related personnel and suspect, for example, relationship, borrowingLoan relationship, lovers' relationship, conjugal relation etc..
Above-mentioned, dynamic relationship network can include the suspect of different time and the relationship change of different related personnel,Such as suspect Xiao Ming is with small red for lovers' relationship before crime, after crime, Xiao Ming with it is small red for conjugal relation, can be closed in dynamicIt is to be shown in network.
It is above-mentioned, different time section is built according to the relation data between the personal perimeter data of suspect and related personnelThe dynamic criminal network of the suspect of point, by the dynamic criminal network would know that the different time node of the suspect from it is differentRelationship and variation between related personnel.
Step S200 extracts the corresponding destination node feature of the suspect in the dynamic criminal network and according to instituteIt states destination node feature and establishes vector.
It is above-mentioned, after the dynamic criminal network is built, based on the dynamic criminal network, extract the corresponding mesh of suspectNode feature is marked, and a vector is formed to the particular network of destination node according to destination node feature.
Above-mentioned, destination node is specific people, can be the related personnel of suspect, or suspect is in itself.
It is above-mentioned, it is to be understood that in mathematics, vector (also referred to as Euclid's vector, geometric vector, vector) refers toAmount with size (magnitude) and direction.It is expressed as line segment with the arrow with can visualizing.Arrow is signified：It representsThe direction of vector；Line segment length：The size of representation vector.
Step S300, according to the destination node feature and the vector structure classification analysis model.
It is above-mentioned, the destination node feature and vector of destination node are calculated, it, should so as to build classification analysis modelModel can be LR models, Multiple regression model.
It is above-mentioned, it is to be understood that logistic, which is returned, is also known as logistic (logistic) regression analysis, is mainly flowingRow disease learn in using more, more commonly used situation is to explore the risk factor of certain disease, certain disease is predicted according to risk factorProbability of generation, etc..For example, wanting to inquire into the risk factor that gastric cancer occurs, two groups of crowds can be selected, one group is gastric cancer group,One group is non-gastric cancer group, and two groups of crowds have different signs and life style etc. certainly.Here dependent variable is just whether gastric cancer,That is "Yes" or "No", is two classified variables, and independent variable can be including it is enough, such as age, gender, eating habit, pylorusPylori (Hp) infection etc..Independent variable is either continuous or classification.Pass through logistic regression analyses, it is possible toIt is the risk factor of gastric cancer substantially to recognize which factor of bottom.In the present embodiment, it is mainly used for forecast analysis to judge and findCorrelative factor.
Step S400 carries out the state supervision to the destination node based on the classification analysis model.
It is above-mentioned, the analysis result of the different time points of suspect is known according to classification analysis model, can according to the result intoRow is to the real-time oversight of the node state of suspect.
The Crime Spreading complex network node analysis method that the present embodiment is provided is started with from destination node, specific with reference to thisDynamic relationship network residing for node, various dimensions structure network characterization, so as to analyze network of the destination node before and after crimeChange in topology obtains the analysis result of suspect so as to fulfill what is analyzed by the network topology change of destination node, pointIt is accurate to analyse result, is that further the analysis to the peripheral information of suspect and investigation bring great side to case investigatorJust.
Embodiment 2：
With reference to Fig. 3, second embodiment of the invention provides a kind of Crime Spreading complex network node analysis method, based on above-mentionedFirst embodiment shown in Fig. 2, the step S300 " according to the destination node and the vector structure classification analysis model ",Including：
Step S310 in the dynamic criminal network, according to the vector of the destination node before and after crime, is obtainedTo the dynamic change characterization vector of the destination node；
It is above-mentioned, the vector of the destination node before and after crime is calculated, so as to obtain variation characteristic vector.
Step S320 builds the crime vector of the suspect based on the dynamic change characterization vector；
It is above-mentioned, according to variation characteristic vector all crime node, that is, criminals are carried out with the structure of crime vector.
Step S330 according to the crime vector, establishes classification analysis model.
After the crime vector is constructed, this base of a fruit regression model of construction logic, wherein, the positive sample in model is from justThe collection of illustrative plates variation of ordinary person's different time.In the present embodiment, it is moved by the vector according to node in dynamic criminal networkState variation characteristic vector, and then crime vector is obtained, and classification analysis model is established according to crime vector, in order to according to this pointThe implementation that alanysis model carries out suspect node state is supervised.
Embodiment 3：
With reference to Fig. 4, third embodiment of the invention provides a kind of Crime Spreading complex network node analysis method, based on above-mentionedSecond embodiment shown in Fig. 3, the step S310 is " in the dynamic criminal network, according to the target before and after crimeThe vector of node obtains the dynamic change characterization vector of the destination node ", including：
In the dynamic criminal network, primary vector is built for destination node described before crime by step S311, forThe destination node structure secondary vector after crime；
Step S312 calculates the difference of the secondary vector and the primary vector, obtains the described of the destination nodeDynamic change characterization vector.
It is above-mentioned, it is V1 for the destination node structure primary vector before crime, for the destination node structure the after crimeTwo vectors are V2.During calculating, pass through the dynamic change characterization vector V3 of V2-V1=V3, as criminal network.
Embodiment 4：
With reference to Fig. 5, fourth embodiment of the invention provides a kind of Crime Spreading complex network node analysis method, based on above-mentionedFirst embodiment shown in Fig. 2, the step S100 " according to the personal perimeter data of suspect and with suspect's relevant peopleThe relation data of member builds dynamic criminal network corresponding with the suspect ", including：
Step S110 builds the network topology of the different time node of the suspect according to the personal perimeter dataFigure；
It is above-mentioned, network topological diagram is built, wherein, the attribute comprising destination node and personal periphery in the network topological diagramData, for example, can include but is not limited to fund, permanent residence, home background, cause situation, age, gender, educational background, disease letterBreath, previous conviction etc..
Step S120, in the network topological diagram, the suspect and suspect's phase to different time nodePass personnel set state relation label；
Above-mentioned, the personal perimeter data based on collection, carrying out labeling to the node in network topological diagram, (all labels are allIt is built by personal perimeter data, concrete operation method is the presetting method of weight.For example, judge rich two generationsStandard is：Father/mother's assets be more than 5,000 ten thousand, frequently borrow money, frequent Capital Flow, vagrant, Fu Erdai, teenager, familyFront yard violence, single-parent family, cause baffle.
It is above-mentioned, carry out labeling to relationship, the content of labeling can include but is not limited to Capital Flow, kinship,Commensurate, same to house, same to historical background (commiting a crime together, same to school), position co-occurrence (co-occurrences such as hotel, Internet bar), telephonic communication etc.Deng.
Step S130 according to the network topological diagram for including the state relation label, builds the suspect notWith the dynamic criminal network of timing node.
According to network topological diagram, the dynamic criminal network of suspect's different time node is built, constructs dynamic crime netNetwork, has also just built the association attributes of people, and then constructs interpersonal relationship, can be understood by correlativity, dividedAnalyse the situation of suspect or destination node.
Embodiment 5：
With reference to Fig. 6, fifth embodiment of the invention provides a kind of Crime Spreading complex network node analysis method, based on above-mentionedFirst embodiment shown in Fig. 2, it is special that the destination node feature includes network statistics feature, network path feature, network subgraphSign, certain service features and the state relation label；
The step S200 " extracts the corresponding destination node feature of the suspect and root in the dynamic criminal networkVector is established according to the destination node feature ", including：
Step S210 according to default node degree, extracts the institute in the dynamic criminal network in the default node degreeNetwork statistics feature, the network path feature, the network Sub-Image Feature, the certain service features and the state is stated to closeIt is label；
It is above-mentioned, it is to be understood that node degree is the subtree number that node possesses；For example, the degree of A is 3.Common data knotStructure includes linear list, queue, stack, tree etc..Tree is n (n>0) (in other words, tree is made of the finite aggregate of a node node).It is known as empty tree as n=0.In any non-empty tree：1st, the node of one and only one root for being known as the tree；2nd, except root knotRemaining node except point can be divided into limited mutually incoherent set, and wherein each set is again in itself one tree, is claimedSubtree for root.This is a recursive definition, i.e., has used tree again in the definition of tree.The definition of tree shows the characteristic of tree,I.e. one tree is made of root node and several stalk trees, and subtree can be made of several smaller subtrees.In treeEach node is the root node of a certain stalk tree in the tree.
Above-mentioned, in the present embodiment, default node degree is set as more than 5 1-5 or, can according to specific operational data amount intoAdjustment of the row to default node degree.
It is above-mentioned, in the dynamic criminal network, for default node degree (from 1 to 5, the maximum number of degrees can change) respectivelyExtract following network statistics feature：Network node number of types, cluster coefficients, average in-degree, averages out average path lengthDegree, degree cluster Relativity of Coefficients, matches together property, is different with property, node betweenness, side betweenness etc. at network resilience.
It is above-mentioned, in the dynamic criminal network, for default node degree (from 1 to 5, the maximum number of degrees can change) respectivelyNetwork path feature is extracted, path is built-up by the label on side.
It is above-mentioned, in the dynamic criminal network, for default node degree (from 1 to 5, the maximum number of degrees can change) respectivelyNetwork Sub-Image Feature is extracted, subgraph is built-up by mulitpath.
It is above-mentioned, in the dynamic criminal network, for default node degree (from 1 to 5, the maximum number of degrees can change), take outTake certain service features：For example, being vagrant's ratio in default node degree, crime ratio etc. in node degree is preset.
It is above-mentioned, in the dynamic criminal network in, extract state relation label as feature, such as age, fund etc..
Step S220, the network statistics feature, the network path feature in the dynamic criminal network, instituteIt states network Sub-Image Feature, the certain service features and the state relation label and establishes the corresponding vector of the destination node.
It is special by the network statistics feature, the network path feature, the network Sub-Image Feature, the specific transactionsSeek peace the feature of state relation label structure, a vector formed to the particular network of destination node, as establish described inThe corresponding vector of destination node.
Embodiment 6：
With reference to Fig. 7, fifth embodiment of the invention provides a kind of Crime Spreading complex network node analysis method, based on above-mentionedFirst embodiment shown in Fig. 2, the step S400 " carry out the state to the destination node based on the classification analysis modelSupervision ", including：
Step S410 analyzes the classification analysis model of the destination node and the vector, obtains the suspectDifferent time node analysis result；
Step S420 carries out the state supervision to the destination node according to the analysis result.
It is above-mentioned, for the vector corresponding to the obtained classification analysis model and destination node, can be analyzedThe analysis result of different time node is obtained afterwards, and the state supervision to the destination node is carried out according to the analysis result.
Step S100 " the relation datas according to the personal perimeter data of suspect and with the suspect related personnelBefore structure dynamic criminal network corresponding with the suspect ", further include：
Step S500 obtains the personal perimeter data of the suspect；Obtain the related personnel and the suspect itBetween relationship relation data.
Above-mentioned, it is personal multi-faceted data to obtain personal perimeter data, such as：Bank data information, is sent at account informationInstitute's information, archive information, information for hospital, cellphone information, mobile phone location information, hotel information, trip information, drugs flow-dataEtc..
Embodiment 7：
With reference to Fig. 8-10, a kind of Crime Spreading complex network node analysis side provided in order to better understand the present inventionMethod provides the 7th embodiment and algorithm in method flow provided by the present invention and flow is illustrated：
Following situation can be seen that by Fig. 8 and Fig. 9：
Before drug abuse, Xiao Ming often occurs with lover inside family, also often strolls store together.Also ex-colleague can be invited to comeFamily is played.And after drug abuse, Xiao Ming and lover needless to say together, often meet with Li Si in hotel instead, but also pass throughOften collect money from the audience to Li Si.It can be evident that the variation of network collection of illustrative plates before and after taking drugs in fact, and this is also effective differentiates generallyPersonnel are changed into the important feature of drug addict.
With reference to figure 10, the carry out latent structure and model construction method in Complex Networks Analysis are as follows：
1st, tectonic network statistical nature.It can include but is not limited to network node number of types, average path length, cluster systemNumber, average in-degree, average out-degree, network resilience, degree cluster Relativity of Coefficients, with property, different with property, node betweenness, side betweenness etc.Deng.In addition to conventional network characterization set forth above, the relevant network characterization of business is may be incorporated into, such as：Xiao Ming's is defaultWhether there are the malicious personnel of suction (dealer) inside the circle of node degree, whether be mostly vagrant inside the default node degree circle of Xiao Ming,Whether the default node degree circle of Xiao Ming is often with vagrant's co-occurrence etc..
2nd, tectonic network route characteristic.And during route characteristic is constructed, structure destination node is first attempted to figureIt composes the path of arbitrary node, and then carries out certain association analysis (AP FP algorithms), analyze some most frequent paths, soAfter further assessed, select some most significant paths as feature.It is above-mentioned, it is described to be characterized as discretized features, i.e.,Whether there is 0/1.
The process of path configuration can include following flow：
Step 1, all nodes are selected as starting point, select all nodes in the default node degree of starting point as knotSpot then looks for shortest path (path represents with relationship, more relationship separated by commas, per once using | separate).It takes drugsBefore, once：Lover, Chang Gongxian is in market, and Chang Gongxian is in family；Old colleague, Chang Gongxian is in family；Two degree：Old colleague, Chang GongxianIn family；Same primary school, shallow bid stablize Capital Flow.After drug abuse, once：Lover；Old colleague, Chang Gongxian is in hotel, shallow bid fundSteady flow；Two degree：Old colleague, Chang Gongxian is in hotel, shallow bid fund steady flow；Same primary school, middle stock gold steady flow.
Step 2, since path is excessive, many relationships are useless relationship in fact on path, so needing for pathAnalysis is associated, finds some subpaths frequently occurred in path.In addition, it is different more to may be based on other methods searchingShortest path feature, for example, path vectorization the methods of.
Step 3, frequent subpath is screened, finally determines available route characteristic for experiment.
Step 4, step 2 and step 3 process is repeated, iterate path optimizing feature.
3rd, tectonic network Sub-Image Feature.Combination (be similar to route characteristic construct) of the subgraph for multipath.But due to moreThe combination in path may some search spaces it is excessive, can search space be reduced by experience, so as to reduce operand, such asOnly look for common classic map：Y type figures, circular chart, Centered Graphs etc..
In addition, network characterization can also include other personal attribute's features, such as personal label and personal attribute etc..
4th, it after it can construct the feature vector of specific node, needs to before drug abuse, the feature vector after drug abuse carries outIt is jointly processed by and (such as subtracting each other), as the last feature (dynamic change for representing feature) of sample, while labeled as positive sample.AndFor the construction of negative sample, the latent structure of negative sample is implemented as by the network of the different time span of normal person.
5th, after sample is got, correlated characteristic is as got, can further be learnt using disaggregated model, fromAnd the information needed.
In addition, with reference to Figure 11, the present invention also provides a kind of Crime Spreading network node analytical equipment, including：Including：StructureModule 10, extraction module 20 and supervision module 30；
The structure module 10, for the personal perimeter data according to suspect and the pass with the suspect related personnelCoefficient is according to structure dynamic criminal network corresponding with the suspect；
The extraction module 20, it is special for extracting the corresponding destination node of the suspect in the dynamic criminal networkIt levies and vector is established according to the destination node feature；
The structure module 30 is additionally operable to according to the destination node feature and the vector structure classification analysis model；
The supervision module, for carrying out the state supervision to the destination node based on the classification analysis model.
In addition, the present invention also provides a kind of user terminal, including memory and processor, the memory is used to storeCrime Spreading network node analysis program, the processor run the Crime Spreading network node analysis program so that the useFamily terminal performs Crime Spreading complex network node analysis method as described above.
In addition, the present invention also provides a kind of computer readable storage medium, stored on the computer readable storage mediumThere is Crime Spreading network node analysis program, realized as above when the Crime Spreading network node analysis program is executed by processorState the Crime Spreading complex network node analysis method.
It should be noted that herein, term " comprising ", "comprising" or its any other variant are intended to non-rowHis property includes, so that process, method, article or system including a series of elements not only include those elements, andAnd it further includes other elements that are not explicitly listed or further includes intrinsic for this process, method, article or system instituteElement.In the absence of more restrictions, the element limited by sentence "including a ...", it is not excluded that including thisAlso there are other identical elements in the process of element, method, article or system.
The embodiments of the present invention are for illustration only, do not represent the quality of embodiment.
Through the above description of the embodiments, those skilled in the art can be understood that above-described embodiment sideMethod can add the mode of required general hardware platform to realize by software, naturally it is also possible to by hardware, but in many casesThe former is more preferably embodiment.Based on such understanding, technical scheme of the present invention substantially in other words does the prior artGoing out the part of contribution can be embodied in the form of software product, which is stored in one as described aboveIn storage medium (such as ROM/RAM, magnetic disc, CD), including some instructions use so that a station terminal equipment (can be mobile phone,Computer, server or network equipment etc.) perform method described in each embodiment of the present invention.
It these are only the preferred embodiment of the present invention, be not intended to limit the scope of the invention, it is every to utilize this hairThe equivalent structure or equivalent flow shift that bright specification and accompanying drawing content are made directly or indirectly is used in other relevant skillsArt field, is included within the scope of the present invention.