A kind of multiclass social application subject information cluster association method based on smart phoneField
The present invention relates to soft projects, data storage, data analysis field, and in particular, to one kind is based on smart phoneMulticlass social application subject information cluster association method.
Background technique
With the fast development of information technology and mobile communication technology, social interaction even works, has traded increasinglyDependent on smart phone.Correspondingly, smart phone information can increasingly react the true social networks of people.In this regard, as public affairsSafety management business department altogether, if it is possible to intuitive relational model is timely and effectively excavated from magnanimity mobile phone, it will sideIt helps excavation criminal gang core member and deep-cuts remaining crime.
Currently, focusing primarily upon mobile phone individual about the excavation of cellphone information, mobile device is relied on as lower peopleSpend it is higher and higher in the case where, the information content of mobile phone individual can be increasing and numerous and jumbled, arranges these information in the whole text and is often difficult toQuality and efficiency is effectively ensured, so the dimension for integrating these information has growing value, statistics social networks are exactlyOne kind of cellphone information statistical dimension.
In the implementation of the present invention, inventor's not strong, number that finds that the prior art at least has a data mining specific aimIt is not high according to relevance, using the defects of experience difference and insufficient.
Summary of the invention
It is an object of the invention to existing a series of problems under background in view of the above technology, propose a kind of based on intelligent handThe multiclass social application subject information cluster association method of machine, is obtained with making up many deficiencies, promotion existing for existing the relevant technologiesThe degree of association for taking cellphone information provides the association between a kind of mobile phone, to reach the mesh for excavating criminal organization's structure and crime process, more powerful technical support is provided for social public security management business.
To realize purpose as described above, the technical solution adopted by the present invention is that: a kind of multiclass based on smart phone is socialUsing subject information cluster association method, specifically include that
A. by smart phone information point application storage；
B. typing reality and virtual portrait ID, bind personage ID and mobile phone；
C. each smart phone classification data d1 is obtained, degree of association set w between weighted calculation mobile phone1；
D. classification data d2, d3 ..., dn, weighted calculation degree of association set w are successively obtained2w3...wn；
E. weighted calculation w1To wnAnd the relation value between each personage ID of Regularization obtains final relationship between codomain [0,1]Set w；
F. personage ID relation line is obtained according to w, the removal degree of association is lower than the association line of lower threshold μ, uses remaining pointPersonage ID discussion on relation is constructed with line.
Further, it in step a, to the operation that cellphone information point application is put in storage, specifically includes:
1, data source is parsed, according to data source, distinguishes Android, the mobile device of ios system, html using analysis programReport, resolves to storage script respectively.It is each application according to using actual conditions at most divide into primary account number, from account, information,The several tables of group.
2, data loading applies off-balancesheet in addition to each, mobile phone hardware, single mobile phone statistical information is calculated and be put in storage.
Further, in stepb, the operation to personage ID typing and binding, specifically includes:
1, mobile phone master known to typing, binds corresponding mobile phone；
2, for dereliction mobile phone, virtual personage ID is created, insert guess or known information, bind mobile phone；
3, for not being put in storage, but the associated personage ID that pays close attention to, virtual portrait ID is established, mobile phone letter is waitedBreath networks.
Further, in step c, include: according to incidence relation, specific steps between classification data d1 acquisition mobile phone
1, for classification data d1_1 (P), d1_2 (C), with the equal inquiry view for establishing CjoinP for condition of cell-phone number(V1), the association mobile phone of the d1_2 class data for inquiring each mobile phone.
2, for classification data d1_1 (P), d1_3 (M), with the equal inquiry view for establishing MjoinP for condition of cell-phone number(V2), the association mobile phone of the d1_3 class data for inquiring each mobile phone.
3, the degree of association of weighted accumulation d1_2 and d1_3 sets up the dictionary (d) that mobile phone id group two-by-two is combined into key assignments, traverses v1And v2 increases the degree of association to dx, wherein λ when mobile phone id meets a line in d (dx)1λ2λ3It is all customized weighted value, associationIt is as follows to spend calculation formula:
λ1·(d1_21)+λ2(d1_22) or λ3·(d1_31)。
Further, in step d, the incidence relation between mobile phone is obtained according to every kind of certain kinds social application, it is specific to walkSuddenly include:
Classification data d2_1, the d2_2 of each social application are obtained, the degree of association respectively applied between mobile phone is calculated,
μ1·((α)+λ1·tanh(d2_1/κ1))+μ2·((β)+λ2·tanh(d2_2/κ2))
Further, in step e, add up more association degree, the specific steps are as follows:
1, the dictionary (d that mobile phone id group two-by-two is combined into key assignments is set upfinal), traverse w1~wn, dictionary is cumulative according to mobile phone idWeight.
2, d is handled using Regularization functionfinalIn all degrees of association: dx.rel=tanh (dx.rel), make the degree of associationCodomain is limited between [0,1].
Further, in step f, all degrees of association and Regularization are integrated, the specific steps are as follows:
Traverse dfinal, minimum threshold μ is set, if dx.rel d is added in point set P in > μxTwo association mobile phonesThickness of two association mobile phone Id as endpoint and the degree of association as line is added in Id, online set L.
Other features and advantages of the present invention will be illustrated in the following description, also, part becomes from specificationIt is clear that understand through the implementation of the invention.
Below by drawings and examples, technical scheme of the present invention will be described in further detail.
Detailed description of the invention
Attached drawing is used to provide further understanding of the present invention, and constitutes part of specification, with reality of the inventionIt applies example to be used to explain the present invention together, not be construed as limiting the invention.In the accompanying drawings:
Fig. 1 is the method signal based on a kind of multiclass social application subject information cluster association method based on smart phoneFigure.
Specific embodiment
Form is described in further detail above content of the invention again by the following examples, but should not manage thisFor solution for the scope of the above subject matter of the present invention is limited to the following embodiments, all technologies realized based on above content of the present invention are equalBelong to the scope of the present invention.
According to embodiments of the present invention, a kind of multiclass social application subject information cluster association based on smart phone is providedMethod can be established effectively using a kind of this multiclass social application subject information cluster association method based on smart phonePersonage's social networks net.
As shown in Figure 1, the schematic diagram of the present embodiment, mainly uses following technical scheme: the multiclass society based on smart phoneSubject information cluster association method of applying is handed over, this method application model includes six parts:
First part is cellphone information point application storage, imports existing iso and Android data source or html formatIt reports in database.
In the database, the data of storage are divided into mobile phone hardware information, application message, application contact, application groupFour major class tables.Such as each browser application has one table of browser application information, and each chat class application has applicationThree message, application contact, application group tables.
Second part is typing personage ID, binds personage ID and mobile phone may include having confirmed that in the personage ID of typingExisting personage and the suspect that do not bring to justice for needing to pay close attention to；
Such as have identical wechat and QQ account logged in the mobile phone A and B of ownership unconfirmed, visual human can be setObject P is associated with A, B.It, can the gradually anti-real person's information for pushing away P after summing up the social circle of P, cell-phone number, permanent residence.
Part III is to obtain each cell phone messaging time length, number and short message record, calculates the call short message association between mobile phoneDegree；
Specifically include the following steps:
1, for mobile phone watch (P), message registration table (C), with the equal inquiry view for establishing CjoinP for condition of cell-phone number(V1), the call for inquiring each mobile phone is associated with mobile phone.The weighted accumulation call degree of association, sets up mobile phone Id group two-by-two and is combined into keyThe dictionary (d) of value traverses v1 and v2, when mobile phone id meets a line in d (dx), increases the degree of association to dx, wherein λ1λ2λ3All it isCustomized weighted value, calculation of relationship degree formula are as follows:
λ1The duration of call+λ2Talk times (1)
2, for mobile phone watch (P), short message record sheet (M), with the equal inquiry view for establishing MjoinP for condition of cell-phone number(V2), the short message for inquiring each mobile phone is associated with mobile phone.The weighted accumulation short message degree of association sets up mobile phone Id group two-by-two and is combined into keyThe dictionary (d) of value traverses v1 and v2, when mobile phone id meets a line in d (dx), increases the degree of association to dx, wherein λ1λ2λ3All it isCustomized weighted value, calculation of relationship degree formula are as follows:
λ3Short message number (2)
Part IV is the address list for obtaining each social application, group, message number, calculates the association respectively applied between mobile phoneDegree；
For example, A mobile phone has QQ account QA, B mobile phone has QQ account QB.Store in A: the two has good friend and chat recordHave 182, there are three common group, there is chat record 15026,55,2896 respectively, then calculation of relationship degree according to:
μ1·((α)+λ1Tanh (single chat number/κ1))+μ2·((β)+λ2Tanh (group chat number/κ2)) (3)
Wherein α β is good friend/common group amendment, is added in chat number weight as long as good friend/in common group.It is falseIf μ1=μ2=0.5 λ1=λ2The κ of=0.8 α=β=0.21=κ2=100, then it is 0.9795 pass that A to B, which just has a degree of association,Connection.
Part V is all kinds of degrees of association of weighting summation, merges association of the association between mobile phone, and will be associated with metricIntegralization is to a limited codomain range；
Part VI is the association screened out lower than minimum threshold, and mobile phone association set is converted to visual character relationFigure, the degree of association represent associated degree of strength between two mobile phones, the thickness of line and point are converted into figure.In order to be closed in simplification figureSystem, the association between mobile phone be it is undirected, the degree of association is derived from bidirectional relationship addition.Public safety can be helped by generating task nexus figureManagement business department expeditiously sorts out character relation from cellphone information, and help understands criminal organization's structure, and excavating remaining crime canEnergy.
Specific steps include: the dictionary (d for setting up mobile phone Id group two-by-two and being combined into key assignmentsfinal), traverse w1~wn, dictionary according toMobile phone Id weight accumulation.D is handled using Regularization functionfinalIn all degrees of association: dx.rel=tanh (dx.rel), make to closeConnection angle value domain is limited between [0,1].I.e. are as follows:
Rfinal=λ tanh (Sum (Rx)) (4)
Such as having multiple applications associated between two mobile phones, degree of association collection is combined into { 0.1,0.3,0.9,0.8 }, setting amplificationMultiple λ=1, ignoring threshold value was 0 (not ignoring), then being 0.97 according to the global association degree for integrating formula this two mobile phones, such asIt is 0.5 that fruit, which ignores threshold value, their global association degree is exactly 0.93.
Correlation model such as figure two is ultimately generated, the higher relationship of the degree of association and the highest mobile phone of participation have bigger halfDiameter.
In conclusion the method based on cellphone information building personage's social networks net, specifically includes that and answers cellphone information pointWith storage；Typing reality and virtual portrait bind personage and mobile phone；Each cell phone messaging time length, number and short message record are obtained, is addedPower calculates the degree of association between mobile phone；Successively obtain social category application buddy list, group-list, chat content, the weighted calculation degree of associationSet；Between relation value between weighted accumulation and each personage of Regularization, final set of relationship w is obtained；Character relation is obtained according to wLine uses mobile phone and task to construct character relation pessimistic concurrency control as line as element in point and w.
Finally, it should be noted that the foregoing is only a preferred embodiment of the present invention, it is not intended to restrict the invention,Although the present invention is described in detail referring to the foregoing embodiments, for those skilled in the art, still may be usedTo modify the technical solutions described in the foregoing embodiments or equivalent replacement of some of the technical features.All within the spirits and principles of the present invention, any modification, equivalent replacement, improvement and so on should be included in of the inventionWithin protection scope.